Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Building a World-Class IT Services Business | Kalyan (QualiZeal) image

Building a World-Class IT Services Business | Kalyan (QualiZeal)

Founder Thesis
Avatar
0 Playsin 57 minutes

In this powerful episode of Founder Thesis, host Akshay Datt speaks with Kalyan Rao Konda, Co - founder of QualiZeal, a quality engineering firm transforming how global companies approach software testing and quality assurance.  

With over two decades of experience in QA, Kalyan shares his journey - from the early days of reactive testing to today’s proactive, “shift-left” quality engineering model. He unpacks what it takes to build a world-class IT services company, grounded in deep client-centricity, employee empowerment, and a willingness to sacrifice short-term wins for long-term trust.  

We also explore how AI is revolutionizing software testing, including: 

👉How QualiZeal’s proprietary “Humanities AI” platform augments engineers (not replaces them) 

👉Solving modern QA challenges like AI hallucinations 

👉Improving productivity by 25–30% with AI co-pilots  

Kalyan candidly breaks down the economics of IT services-from pricing, utilization, and accelerators to gross margins and client segmentation. He explains why value buyers win, how accelerators and COEs give firms a strategic edge, and how to stand out in a crowded, commoditized market. 

For aspiring entrepreneurs, Kalyan offers hard-won insights on: 

👉Founding a company at 45 

👉Sales strategies that emphasize relationships (“People buy from people”) 

👉Navigating legal challenges while scaling 

👉Finding your niche and staying the course  

This is a must-watch for anyone building in quality engineering, IT services, AI-powered testing, or enterprise SaaS.  

#QualityEngineering #SoftwareTesting #AIinQA #ShiftLeftTesting #QAPlatform #FounderThesis #ITServicesBusiness #EnterpriseSaaS #StartupJourney #SoftwareQuality #AIinSoftwareTesting #ClientCentricity #HumanitiesAI #BusinessOfQA #Entrepreneurship #Founderthesis

Disclaimer: The views expressed are those of the speaker, not necessarily the channel.

Recommended
Transcript

Key Questions for Business Success

00:00:00
Speaker
We are willing to lose in front of clients for the clients to win. ah Three important things are business too. One is actually, is there something that you are deeply passionate about? That's one.
00:00:12
Speaker
Another thing is that, can you do this thing better than anybody else? Right, that's another one. The third thing is that if you do those two and other people that are willing to pay for it.

Podcast Promotion

00:00:27
Speaker
Are Indian companies more quality conscious or Chinese companies?
00:00:33
Speaker
If you enjoy such deep dive conversations with real builders and founders, then like share and subscribe to the Founder Thesis Podcast.

Kalyan's Career Journey

00:00:42
Speaker
Kalyan, please,
00:00:46
Speaker
can you take us through a little bit of your background? What got you here? Sure, ah sure, actually, yeah. Thank you so much ah for having me on the show. Really appreciate it and they're looking forward to this com conversation.
00:01:00
Speaker
So I have i started my career in and software testing. That was a back in 1998. And i I started my um journey with the my corporate journey with a company named Bond.
00:01:20
Speaker
Bon was an ERP company at that point in time so I actually started working in the Bon product testing group and and actually really enjoyed that stint Bon had really solid solid
00:01:43
Speaker
fundamentally it was a great company and and actually it focused a lot on software testing as a as a function um ah so i really wanted to experience the services side of it and actually i went and joined over chusa and then after ah having spent ah you know two years there and then actually ah Then I joined a company named Applabs.
00:02:15
Speaker
At that point in time, Applabs was a relatively small company compared to Wachusa.

Founding Quality Zeal

00:02:23
Speaker
And the Applabs, what attracted me to Applabs was, Applabs was just focused on providing software testing services.
00:02:34
Speaker
and Unlike Wachusa, Wachusa was actually doing all the all the services, I mean, the complete, complete spread of IT services and and actually testing services being one of them.
00:02:48
Speaker
ah but But then App Labs had a focus on software testing services and they were only doing software testing services because i was anyway very keen on building my career in software testing.
00:03:04
Speaker
It made perfect perfect sense for me to join App Labs and ah joined App Labs and and very quickly i was in the management team of app labs and and then i was in app labs for almost 10 years and the entire a organization just grew in front of me and then actually decided to join a company named signity signity again at that point in time was a very
00:03:42
Speaker
even smaller than Apple apps when I join significant was relatively like they just had about less than 20 people and I've joined their management team management team on day one and I ah had actually stayed in Signity and all the way till end of 2021 and I actually got an opportunity to focus more on the business development side, essentially building the the sales organization to open new logos and also an organization to basically to grow ah the existing accounts.
00:04:24
Speaker
And I was in Signity almost till the end of 2021. And and And right after that, the the quality zeal was born and then you know about seven people that came together and I was i was one of them.
00:04:45
Speaker
And we we all had one thing in common that is ah a passion for software testing. I mean, you know by then, as you know, the industry,
00:05:00
Speaker
ah changes the jargons. Now, you know, once I decayed out, so now it become a quality engineering. So what really drew all of us together is a passion for quality engineering.
00:05:12
Speaker
And then and they' actually we got together and started this in the the in the last quarter of 2021. Passion for quality engineering does not necessarily imply starting your own venture.
00:05:30
Speaker
Was there a business opportunity?

Building a World-Class Organization

00:05:33
Speaker
What what was the business case for it? Yeah. Yeah. that's ah You are absolutely you're absolutely right, Akshay. And just because you are you're passionate about something it doesn't mean that you you jump into it as a business because, ah ah you know, I read a lot of books. so One of the books that I really like is ah Good to Great from from and Jim Collins.
00:06:04
Speaker
And actually he talks about ah three important things to to really start a business. to One is actually, is this is there something that that that you are deeply passionate about? That's one.
00:06:19
Speaker
Another thing is that, can you do this thing? um better than anybody else right that's another one the third thing is that if you do those two and are other people that are willing to pay for it right i mean these these three things are very very important and i think at that point in time all ah all seven of us that came together, except one person, everybody else, ah we were all first-time entrepreneurs.
00:06:54
Speaker
And ah so, of course, you know, entrepreneurship was at the back of our mind. But these are the three checks ah that we had done. and But you know you were already running Signity.
00:07:12
Speaker
You were the president there. Essentially, what was it that that you couldn't have done at Signity itself. If you were amazing at delivering quality engineering, you could have done that within Signity.
00:07:27
Speaker
So what was the reason to start? Yeah, so one one thing is that I i sort of, ah one, there was that the that each of having something on your um anur won was was always there.
00:07:52
Speaker
But however, I would say that ah courage was lacking. Courage to take the first step was ah was ah clearly lacking. And ah that That was number one. I think ah number two was also that if if you look at ah ah the way things things happened in my career, ah that actually that but when I joined Apple AppSyft, it was fairly small. I really enjoyed so
00:08:23
Speaker
the the building it building it all the way up to... um ah being part of the journey and same with the uh signity right and and actually when when you get to a certain size yeah ah then actually the the administrative work also takes over um doing some real good high end quality work doesn't mean that administrative work is not high end quality work, but then, you know, personally,
00:08:56
Speaker
um ah you know, there is a pleasure of ah building things from the ground up that that always fascinated me and and also interested, very much interested in doing it.
00:09:12
Speaker
And also having having seen two organizations up close um ah you know, building something in the quality engineering space, I think that also gave, would say both, the experience was always there. What was missing was the courage to do. I think, you know, it took me some time to become courageous. And also I would, we were, it was also very clear that the entrepreneurship
00:09:51
Speaker
um the uh actually the chances of uh being successful is very high that that actually if you can build a solid team on day one and a team that also cares about the collective vision and also so the team that is actually capable of executing on that vision so so yeah i think those those those are the things that came together ah to to really start something actually was it also a wealth creation incentive
00:10:30
Speaker
that that you felt that you can build a business and it can be a large profitable business and large wealth creation opportunity. Yeah, absolutely. That that that was also that was also on the mind. But but i think what was more but was more important is that um well
00:10:56
Speaker
you know ah The challenge of creating something smoking from the ground up and and actually ah you know not proving it to a single individual, but proving it to the world that we can do that too.
00:11:11
Speaker
I think that was um the that was the primary driver. And I would say that, yes, if we can if you can do that successfully, um then the wealth creation is something is of a solid, but important by-product.

Industry Experience and Founding Teams

00:11:31
Speaker
Got it, understood. The seven of you who came together, was it engineered or was it a lucky ah collection of seven? like like I mean, was it engineered in the sense that did you feel that, okay, I need one finance need one this guy, I need one that guy. And so therefore, given that I need these people, these roles to be filled, these are the best people to fill those roles.
00:11:57
Speaker
Was that how it happened? was it that ah a group of seven friends got together and started quality uh not really that's a great question it was ah it was more of um it was more of uh like you know who can be part of this venture um and uh and then actually uh having some um you know offline talks about it and uh and then actually coming together. Yes, we we want to do it. Then actually we went about identifying who do what.
00:12:35
Speaker
So I would say it's less off it's less of ah identifying what needs to be done and then fitting people into it. It's more of getting the right-minded people sharing the common vision, then figuring out who will do what.
00:12:55
Speaker
Would you do the same thing again? Like a friendship over role fitment? yeah Yeah, absolutely. Any day. I guess ah i guess what what also... I mean, there were there were some things. And I would say probably two months or three months into the into the journey.
00:13:18
Speaker
clearly figured out that there were some parts of the organization that actually, you know, none of us really had had skills.
00:13:29
Speaker
I mean, you know, to run it world class. I mean, we all have skills to to run it at a surface level, but but we are here to create a world class company. So we clearly knew that we didn't have it.
00:13:43
Speaker
And ah so then we had to go acquire those skills and then in some sense, you know, create enthusiasm in ourselves, some sort of ah faking that enthusiasm to acquire those skills.
00:14:03
Speaker
and Because that is also, that's also very

Client-Centric Philosophy

00:14:07
Speaker
important. And then, you know, going and acquiring those skills and running it. Because we, at that point... What's an example of...
00:14:15
Speaker
I would say one of the examples was ah finance. Okay, interesting. Very interesting. Okay. So, you know, what is the advantage someone like you all ah who have seen the industry have over somebody who's very young and starting a business?
00:14:37
Speaker
Like, what's the advantage that you get at your stage of life when you start a business? I would... I would say that, yeah I mean, there are both the advantages and disadvantages, you know, at every at every stage of life, depending on um what you are what you're dealing with.
00:15:00
Speaker
And i would say that, especially ah on the on the and the business side of things, we had enough experience and expertise ah in terms of,
00:15:14
Speaker
convincing a client to do give business to us and also the necessary expertise to be able to deliver compared to any other company that is out there.
00:15:29
Speaker
So that was always there on and day one. you know For example, ah like and I could be sitting across the ah table with the CIO of a Fortune company and uh and actually you know and doesn't really have to uh rely on the uh the the the strength of the the quality zeal i mean you know there's also things that we have done in the past and that will come uh very handy actually and that was that was definitely there and also um i would say another thing by then i don't think uh
00:16:10
Speaker
yeah none of us are completely financially independent per se, ah but at the same time, ah accumulated enough and so that we can actually, built ourselves, I would say enough runway, right? to ah To last even without, you know, single dollar in revenue.
00:16:34
Speaker
i mean, you know, that runway we had was ah almost about 18 to 24 months of runway was created on day one. Okay. as Essentially, we had 18 to 24 months to fail. So we are not looking at, you know, six to 12 months to fail.
00:16:53
Speaker
Okay. um So these are like, you know, tactical advantages that it is easier to sell and the the ability to deliver like like You don't need to experiment and learn how to deliver. You already know how to deliver.
00:17:11
Speaker
What about from a mindset perspective? Like what is the mindset which a 45 year old founder, first time founder brings to the table, which 25 year old founder, you that you would advise a 25 year old founder to have, you know, something which as 45 year old founder, you bring to the table, first time founder.

Quality Engineering and IT Evolution

00:17:35
Speaker
I would say that, um would say that ah you know, 45 year old founder might be approaching things with lot more pessimism and and bucket yeah being realistic about things. And probably the 25 year founder may be approaching it from from a lot of optimism, I guess, from um just from a mindset up point of view, because because all the time when making business plans, when thinking about things, all we were thinking about is that in what different ways it can fail, as opposed to in in in what all spectacular ways it can grow. So, ah right? so
00:18:21
Speaker
and And the good thing is that, ah um the team ah is ah is basically a good mix of optimist, pessimist and realistic people. it's ah It's a good combination. So they there was always that um multi multiple views and ah you um often, especially in the early days, contradicting views.
00:18:48
Speaker
And so that that helped a lot in and balancing things, Akshay. I think there's one more thing which you mentioned, which I think really the massive difference in mindset is you said we are here to build a world class organization.
00:19:09
Speaker
I don't think a 25 year old founder understands that kind of mindset, right? Like to build a world class organization, what does it mean? like Can you talk to me a bit about that?
00:19:21
Speaker
Yeah, ah so like ah like for example, from day one, be we were very clear um that actually we we want to offer the the best possible quality engineering service that is out there.
00:19:39
Speaker
And also, um you know we were client-centric organization from day one. and and one of the principle uh that that actually we wrote it down um and even now we we live by is that we are willing to lose in the short term ah to gain in the long term right and and we are willing to um you know, lose in front of clients for the clients to win, right?
00:20:12
Speaker
So the client centricity was very, very important to us because, and also we are looking at more of a lifetime value of the client as opposed to the initial project and the project after.
00:20:28
Speaker
the that was one which was which was worked as i think a lot of organizations uh sort of struggle with that concept in their minds but i guess i guess we never had any doubt about um the philosophy of willing to lose in the short term versus gain in the long run that was one and the another one was also that we know that we we are building an organization um that has to ah really scale and the scaling is only possible um if if we can uh hire the right talent and actually keep them right and people may be able to hire if they can't keep then then actually we are going to have a lot of these revolving door issues so that was also very important for us that now that actually
00:21:27
Speaker
you know the e the whole employee centricity and the client centricity were the two principles and you know more or less ah both given equal priority actually if i understand correctly like employee centricity and client centricity is what leads to building a world-class organization and of course the principle of willing to lose in the short term ah give me an example of willing to lose in the short term willing to lose in front of the client for the client to win but both of these principles are very intricate yeah so so i would say that ah willing to lose in the short term i mean you know the many it services companies that are out there and initially in a lot of companies a lot of enterprises they have they have options and and they would they would like to

Client Relationships and Empathy

00:22:27
Speaker
you know their procurement departments are looking at a try and buy approach right they they just want to get a feel for what you can do ah before ah before they go all out with you so in those kind of situations i i had seen a organizations um right where they want to they want to calculate margins on day one and and actually very short-sighted about ah making money on on every single uh uh project that they do uh but we we sort of ah did not approach it like that we we approached it more of we took a top-down view of things okay i mean you know
00:23:11
Speaker
this is an enterprise that that we're dealing with what is their outsourcing budget and over a period of if we do well over a period of two three years can we actually grab 30 of that budget or 40 40 of that budget or 70 of that budget and approach it more from that way as opposed to ah the initial 100k project how much money that we're going to make action okay okay interesting um and uh willing to lose in front of clients for the clients to win well i would say that there were the there were there were many situations um so wherein actually the the
00:23:58
Speaker
the clients um are uh are actually have have that point of view and not necessarily that point of view is always always correct.
00:24:12
Speaker
And so so so, which means that we are not here to correct them on day one. We are here to work with them, understand their processes, culture.
00:24:25
Speaker
And so ah so like, you know, we are not trying to correct them on day one first. And so even if it feels off, um in the initial days. So in some sense that, um you know, we we focus the first initial, the months and quarters of the engagement to to really figure out, um right? At that point in time, sometime from a logic sense, it may not make make right sense whether what they're asking us to do is the right thing or not. But then, you know, since we take a long-term view, and so it is okay to,
00:25:05
Speaker
ah to basically to let them have their way um and and As we are trying to navigate and figure out the overall equation
00:25:18
Speaker
Okay um When you are doing multi-million dollar deals with companies
00:25:26
Speaker
Don't they expect you to give thought leadership? And isn't giving thought leadership the same thing as telling them that you are doing things the wrong way? Yes, that is that is absolutely that's absolutely right. Even we thrive on thought leadership.
00:25:46
Speaker
ah But again, unless and until you invest um yeah the the amount of time, and really understand their world.
00:25:58
Speaker
um and then And then actually, even the best advice also will not resonate unless until it is provided at the right time with the right understanding.
00:26:09
Speaker
and And actually, so that is more important. Sometimes the same advice
00:26:21
Speaker
you know, two different people giving it, but one gets accepted, another one doesn't get accepted. it There is a reason why it doesn't get accepted. Maybe it's the way that ah some of the softer aspects of the thought leaderships are equally important.

Shift-Left Approach in Quality Engineering

00:26:35
Speaker
i What influences a client to choose one vendor over another? And I mean, I'm sure the answer here is like a weighted average.
00:26:49
Speaker
You know, so it could be price, it could be pedigree, it could be the quality of the salesperson who's talking to the client, things like those.
00:26:59
Speaker
And these would have different ways, like some would be very important, some would be somewhat somewhat important. So, you know, I just want to hear your take on this, like, you know, what influences client's decision to choose one vendor over another?
00:27:13
Speaker
So I would say actually if, especially in the initial days, and now we are totally different, is that if if somebody has to um the you know take all the parameters, um like the parameters could be you know the domain experience that we have, it could be they the, you know since how many years the company had been founded,
00:27:42
Speaker
and the size of the organization, and how many client referrals that we have, and things like that, we we have no chance of winning any client. mean, let's be honest.
00:27:54
Speaker
So because we would fail on each one of those parameters, or we would rank low compared to compared to other established companies that are out there.
00:28:06
Speaker
um So the fundamental principle is that people buy from people that they like
00:28:15
Speaker
yeah yeah considering everything else is the same and even if it is different and the second one is that they should ah they should feel like no here is a company that they can work with even during a difficult times.
00:28:33
Speaker
um and When things are going well, it's ah it's easy to work with, it's easy to work and the relationship gets really tested during the difficult times. um And I think that made a huge difference, ah that kind of mindset made a huge difference ah to us, especially
00:28:51
Speaker
in in onboarding some of the Fortune clients actually. So you you said people buy from people. so this is like a So, if you understand this, then you'll be able to win deals. right Now, what what does it mean to understand this, that people buy from people? what How does it influence your behavior when you're in front of a client?
00:29:16
Speaker
So, ah like what what we had seen in in many in many client organizations is that um there ah there is ah There is a level of service quality that is ah that is expected from um everybody that there that they're working with. And actually, ah some of the some of the projects that we run are also mission critical. so you've um And you know people people's so jobs are at stake if some of these projects ah don't go as well as they are planned.
00:29:53
Speaker
And, but however, ah they also want to work with people that that are ah truly understands what they are dealing with on a day-to-day basis and are actually willing to ah change priorities and willing to do things um that are that are important to them. Not everything can be spelled out in contract. In fact,
00:30:21
Speaker
everything that we do probably 50 the things that we do cannot be really articulated into a contract so that's what i mean by people buy from people i would say that ah you know in in many situations the way that we a problem is really an opportunity for us to build a relationship with a client ah so the the the issues and uh are there in every every project and that's part of a day-to-day life. And that is the reason why why so many executives are there and solving them.
00:30:57
Speaker
But then how are we looking at that issue risk? Are we looking at this as an opportunity just to solve that? Or are we actually using it to build a relationship and but that will last longer?
00:31:14
Speaker
I think that's ah that's the big difference in how we approach things, Akshay. You're saying that how it would influence your behavior is you would be more willing to drop everything and cater to the client requirement or like go for a meeting or whatever. Show them that you're willing to go the extra mile for them.
00:31:38
Speaker
but that Yeah, show them that we care and and actually spend lot of time and energy in... in understanding their point of view.
00:31:49
Speaker
And i would say of being, showing that empathy, what and really thinking from their point of view, I think goes ah goes a long way. Not only you get the problem solved, but then you also ah build the ah build a platform, build the trust that is ah ah needed ah to continue to do business with these folks in the long run. I mean, we yeah we had we had the seen situation um many situations in the company where actually no ah typical edition maker um you know are changing their jobs like once every three four five years and and we get a tremendous amount of businesses when they change their companies they landed somewhere else and and actually they want to they want to prove themselves and they want to get re
00:32:45
Speaker
sort of established in the new organization and uh and one of the thing that they have uh in their mind is uh actually i'm going to bring quality ze here and i think they will make me successful uh so that's the kind of mindset that we're built with these executives okay amazing amazing amazing um okay Understood.
00:33:08
Speaker
um Let's talk a little bit about the industry that you operate in. Quality engineering, which earlier was software testing. So, you know, I'm not an engineer.
00:33:19
Speaker
I cannot write a single line of code. For someone like me, can you help me understand what is the software testing slash quality engineering? engineering
00:33:28
Speaker
Yeah. So, a lot of enterprises like banks, insurance companies, airlines, cruise lines, manufacturing companies, all of them use software to to run their data business operations.
00:33:47
Speaker
and general Like I guess it was probably 10 years before there was this slogan that every company is an IT company. And irrespective of the business that you are in, you you need to have a very robust ah IT systems to be in place to to run it effectively, irrespective the industry that you are in.
00:34:09
Speaker
and And actually, since IT systems have become ah such an integral part of ah running the day-to-day business operations, and in in many situations like right now, in many industries, IT systems is also a true differentiator for the company. If you could have a better IT system ah to sell,
00:34:33
Speaker
ah deliver and collect revenue. and And actually that you you can be you can be ahead of other other companies in the spain space and it could be an insurance company. As an insurance company that actually if you can if we can provide a code with a high amount of accuracy um in no time and also offer your service through multiple channels,
00:35:02
Speaker
and so that can be a true differentiator i mean now now it's nothing to do with your ability to uh to to basically quantify the risk and also uh also come up with the insurance amount which is which is supposed to be the the core bed and bread of insurance uh company but but then actually if you're using it systems to all to do all that very effectively that can become a differentiator and ah so but however those IT systems ah can i also ah can also become a dragger for your business, especially you for if those IT systems are not functioning the way that they're supposed to do, if those IT systems have security holes, if those IT systems are unable to scale from a from a performance point of view as you as you bring in
00:36:02
Speaker
ah more and more clients and some of these are real issues right so ah so there are ah so like ah for example having a having a reservation system and that does the pricing not accurately can lead to
00:36:24
Speaker
loss of sale and having having a desk management, service desk management, and actually not being able to resolve ah some of these customer issues quickly can lead to client attrition.
00:36:39
Speaker
So I would say that the quality engineering is ah is is basically he advocating ah for for the end user and within the IT t team.
00:36:51
Speaker
and ensuring that ah we are actually helping to build systems that work, that function correctly. And ah being part of the IT systems is the core work that we do. Basically think of think of the quality management, the quality engineering as a as the the biggest ah advocate of the end user within the IT project team.
00:37:20
Speaker
okay oh So but the the roots of this is testing, right? Yes. the The roots of quality engineering is testing. So essentially it means testing the solution from the shoes of the user, like putting yourself in the shoes of the user and then testing the solution to make sure that the user's experience is not broken at any stage.
00:37:49
Speaker
Absolutely. I mean, that's the, I would say that that that's the basic foundation. I mean, it can get a lot more deeper. ah Like, you know, if that's the end objective, actually how do we ensure this ah from day one of the IT project?
00:38:06
Speaker
That means, you know, how do we closely work with the design team in ensuring that, you ah the the the design is being ah built in a way that it ultimately satisfies the um the end user needs.
00:38:20
Speaker
And then how do we how do we partner with the software development teams very early on, like writing efficient code and and designing the user interfaces that are that are easy to navigate and also with less number of clicks and with less number of screens you can get your a work done and and things like that so so now if you if you look at ah the the ah quality engineering

Agile Methodologies in IT

00:38:50
Speaker
right now is is approached more of a
00:38:56
Speaker
ah sort of a co-builder ah right ah working along with the development teams uh from day one and and actually proactively uh ensuring quality as opposed to reacting it towards the end of the life cycle and trying to find as much as you can right so the the approach is now more more of you know finding those uh defects early and when they're actually young
00:39:30
Speaker
To me, this sounds a lot like bringing design thinking into creating software products. Is that what it is? so Yeah, in some ways, in some ways ah but that's what it is.
00:39:43
Speaker
And because also that typical IT lifecycle, when it was ah especially pre-agile days, there was... a you know, less focus.
00:39:59
Speaker
mean, probably there was there was a good amount of thinking around um and around what the user wants in the initial phases. But then after that initial phase, which is basically putting together the business case and functional specifications,
00:40:18
Speaker
ah and then actually for a good amount of time, for many months and years, depending on the complex of the systems, nobody's really thinking about the user. Everybody was ah thinking of trying to figure out the programming language and databases, architecture and writing efficient code and very little to no thinking about but how exactly it is going to be used in the field.
00:40:46
Speaker
Right. I think that contributed to that contributed to a lot of issues. And then and then actually projects not getting complete and then even if they're complete they're not not even close to ah what the business teams so wanted when you know that was the whole genesis uh for this ah entire agile where it was more of how can we uh how can we sort of this eat this elephant in pieces um so that at the end of that actually we produce ah
00:41:22
Speaker
software faster and also put a version that can be actually be used by the ah used by the end users um more frequently and get that feedback and incorporate that feedback into the overall life cycle um right am i i guess ah that kind of it was a similar thinking even within software testing as well so so how do we ah how do we get that um
00:41:50
Speaker
user advocacy end user advocacy into the it project teams um earlier than later okay ah so you're saying that the way software was built earlier was the way let's say a highway is built like there will be a plan drawn out that this is how the highways to be built and then all the different stakeholders will go ahead and start executing it. Like someone will go acquire land, someone will start playing the bricks, et cetera, et cetera. Someone will start procuring, et cetera.
00:42:23
Speaker
So that is how software used to be made. ah And then Agile changed that. What does Agile do? What is the Agile philosophy? So I would say that staying very close with the the end user all the time.
00:42:43
Speaker
and then actually and building something building things that are useful um ah in a in a short cycle of time and actually getting that continuous validation ah from the end users is essentially the whole philosophy of agile and also that from an end user point of view also that many times if i would ask end users on the kind of systems that they really want they know what they want ah but they really don't know till they see it and use it right and many times so they are also figuring out what is useful them to them what is needed based on what you put in front of them right and
00:43:42
Speaker
Once they see something, it is easy for them to tell you what is working, what is not working, what needs to be changed. I think the Azure really brought the the IT teams and business teams as ah together.
00:43:55
Speaker
And ah so that way they are producing stuff, um ah you know, accurately. i mean, I would say more of ah ah what is useful.
00:44:07
Speaker
um And because of this entire model, ah It also allows them, it it gives a flexibility to change things. um The earlier models didn't give that flexibility. Earlier models i expected or assumed that the end users know what they want on day one and they know it 100% and with the no room for changes.
00:44:31
Speaker
Okay, that's so true. Yeah. um So, and Agile is essentially like a
00:44:40
Speaker
Software production methodology?
00:44:44
Speaker
Correct. OK, OK, OK. It is like some series of steps and some best practices of how to produce a software. OK, understood.
00:44:54
Speaker
OK, so now what you're talking about as what the quality team does to me sounds a lot like what, say, a product manager does. like a product manager is the advocate of the consumer.
00:45:07
Speaker
and Like say a company like Zomato would have a product manager who would constantly be thinking in terms of how do I make it easier for Zomato user to order his favorite dish with the minimum amount of friction.
00:45:21
Speaker
um So how is quality engineering different from what a product manager does? the the The product manager is ah is basically laying out the the overall the business ah ah specifications on what the but the software should do um right but then actually the ah quality team is the one that is <unk>s really checking whether it is doing what it is supposed to do that's number one more importantly it is also checking whether the software is doing what it is not supposed to do as well so
00:46:02
Speaker
There are many times, there you know for example, joe you know the authorization is a big part of a lot of IT systems where but there are serious privacy issues.
00:46:18
Speaker
So like Kalyan, who doesn't have authorization to see something, ah should not be able to see, right? And he's not authorized to do something, should not be able to do it.
00:46:29
Speaker
so So essentially the the software test team and the quality team is ah is ensuring what are the product manager envisioned from an end user point of view, ah the system is is actually in line with those specifications.
00:46:48
Speaker
Okay.
00:46:51
Speaker
So this sounds like what testing is, right? Essentially once the product is built, before it goes to a product manager, there's a level of testing done. You spoke of the new new way of doing it ah where you are involved right from the design stage. So what are those the additional things which the quality team does then? Yeah, I would say that if if you're actually testing your product almost towards the end, and and actually one could do that without having serious understanding of, I would say, in-depth understanding of the architecture, design,
00:47:28
Speaker
quo as long as also a lot of the complexities involved in the software processes. But however, um by then, as we already talked, it is a is are too late in in in finding defects. And also it's very expensive to fix those defects.
00:47:51
Speaker
So the whole, ah then the new ah new paradigm is all about shift left right basically you are shifting testing activities um early in the life cycle of the IT project as compared to doing it on the the right um on the right side of this IT t project life cycle so shifting left actually has a lot of uh a lot of issues like now as a as a as a quality engineer within the project you have to have a ah real solid understanding of the design and the architecture because ah um you know in a true shift left approach you are not only you you don't have a user interface to test you are actually looking at looking at block diagrams you are looking at flowcharts and then trying to figure out in what ways it can fail right and you are actually looking at
00:48:57
Speaker
these are these documents the user studies that are being written and so you are actually trying to find defects even without a single line of code being written that requires different skills that requires through the business skills architecture skills design skills and then then as the code is being built up and even sometimes you don't even have a user interface you should still be able to find defects within the code and then there are a lot of sophistication that emerged in this area where actually you have to as a test engineers you you have to write code to test code right essentially somebody is writing code to build application now you are writing code ah to test the code that is being used to build applications and and then and then you know gradually it progresses to
00:49:46
Speaker
more of what the user says, really the screens and then that kind of a

Role of User Stories in Development

00:49:52
Speaker
thing. So, mean, the shift left itself is actually the ah kind of skills that are needed and the attitude that is needed um yeah is completely different from doing things in the end, actually. And that presents ah an enormous opportunity. And why all of this is needed? All of this is needed because,
00:50:13
Speaker
ah you you really want to build the system right from the from the ground up as opposed to, um you know, finding them later in the life cycle and then fixing it. This way, I think the cost of quality, you know, all the activities that you do to ensure the the overall quality of the product is what it is.
00:50:36
Speaker
Doing this way, i think it saves a huge amount of time and also costs for the enterprises. mean, that is the reason why it is being done at a, a very high level. Okay, ah interesting. Shift left as a philosophy is very interesting. I'll go into that in detail a little later.
00:50:54
Speaker
um So, from what understand, the way you produce software, and I'm just basing this on what examples you were giving me now. So it starts with some flowcharts and then there are user stories. So what's a user story? And what would, like, just take me through these steps and then there is code.
00:51:14
Speaker
ah then there is a front-end user interface and that's how you produce software am I right in that basic understanding but that that is correct I mean you know it it all the it all the typical IT project it all starts with the yeah yeah say that somebody is ah designing a reservation system right for airlines right then they would they would basically go about you know how does a reservation system looks like.
00:51:44
Speaker
And typically it all starts with Word document, right? And probably with a few screens on what it should look like. And then you you go about explaining the functionality of each screen and you go about explaining the functionality each module.
00:52:03
Speaker
And then actually, then the people are actually interacting expected to interact with these systems there are all kinds of users and then and then their their user those are basically called user stories right and if ah like what is the uh how do I go about making a reservation and then explaining them in steps that's one user story there could be another user story to cancel a reservation if I if I want to cancel an existing reservation how do I go there is a there is a use case for
00:52:35
Speaker
modifying a reservation the existing reservation how do you go about doing it these are all um user stories and then you actually don't go ah build a system with all these user stories in one step you you would actually then and all these user stories are written then then the it teams takes this user story as an input and then and then probably they would first start building ah making a reservation and then as as they're done with it then they would actually go about making this ah sort of changes to the reservations. Those user stories evolve and they get get implemented in in software one after the other actually.
00:53:14
Speaker
OK, OK, OK. ah And these user stories, this is what is like a business requirement document which includes of these user stories? This is a business requirement, but a lot more, I would say that a lot more concise. and And the the user story also is supposed to be standalone, that it can actually deliver something useful for our end user. ah right When that I'm done with the done with the user story, means that now the users can go and do something about They're able to complete a task.
00:53:53
Speaker
OK. So the the first role of the quality engineer is to actually audit this Word document, which has the user stories to make sure that ah the vision so it all starts with the vision of a product guy like a product manager or whosoever uh the the business guy so his vision will come in probably like a smaller document like an executive summary kind of a document which is the the business requirement document from what i understand yeah the business case high level business requirements okay okay okay then this is converted by somebody in the tech team into these user stories uh
00:54:31
Speaker
And that is then audited by the quality person. And then these user stories, one by one, code is written for them and the front end is made. ah And so again, the quality person will audit the code, audit the front end, and make sure that it is aligned with the user stories. So so this is how it works.
00:54:50
Speaker
that that that That's how. and then And then actually, just to sort of not to complicate, but then, you know, typical IT teams, there are ah let us say if you're building a very very complex system then then there are multiple teams that are working on uh right and there is a there's one team that is working on user story one probably there is another team that is working on user story two it's very typical to have like uh you know five
00:55:21
Speaker
five teams that are working parallel, 10 teams, 15 teams, 20 teams. and ah And they are there is some orchestration and coordination that is needed among the teams that that we are actually ah building things. Ultimately, they had to work together at some point in time down the road.
00:55:43
Speaker
So, and thinking carefully about how they get integrated and when to integrate them, all of that also consumes a lot of brain power of the team's action.
00:56:00
Speaker
So each team would have ah quality engineer and then there would be like a quality manager who would look at the orchestration and making sure these things are working together. Absolutely. there Typically, the the the quality engineers are embedded as part of each of these ah each of these teams. i and they're Now they're calling it as a pod.
00:56:21
Speaker
pd and each part has a certain roles. There is a, there is a scrum master and there is a product manager, developers and test engineers, UI UX, database engineer, depending on the kind of system that you are building. And so some, a team that is a,
00:56:47
Speaker
sort of self-managed team and and and and has everything that they need to be able to ah produce something concrete at the end of each sprint.

Outsourcing Quality Engineering

00:57:02
Speaker
What does the Scrum Master do? g Basically, the Scrum Master in a typical... Like a think of Scrum Master as more of a project manager.
00:57:14
Speaker
Okay. ah good But then the then actually the the interactions are almost daily. Teams are sitting next to each other and they're meeting multiple times a day.
00:57:28
Speaker
But formally, at least they're meeting twice a day. Okay, okay, okay. Understood. So that that is how software is built through pods, each pod having one quality person.
00:57:39
Speaker
Why is the... why Why is there the need for a standalone quality business then? Because from what I understand then, this pod itself could be like an Infosys providing a pod.
00:57:52
Speaker
So when Infosys gives a pod, they would already give a quality person within that pod. Why is there a standalone quality business existing? And if it is being done in-house, if companies are building their in-house pods by hiring people, then they would again hire a quality engineer in-house.
00:58:09
Speaker
So what is the need for what you offer sure yeah i think there are two questions there the one is actually if the teams are building it in-house why not they have it themselves that's a great question so typically let us say that you know somebody is embarking on oh on producing a world-class product right and and actually so so there are there are some that are very core
00:58:39
Speaker
uh core activities that are non-core activities the core activity is basically the product manager thinking of about the product and addressing a gap that currently exists in the marketplace that you could exploit and also also then actually having you know architect ah to to work that this entire it system in a way that it functions well it can scale and it is also no the security is built in and things like that there are some activities that are very core to product building your product right oh ah if ah if i was the one that is embarking and something like that i would actually keep some of those core things within my organization and then actually
00:59:33
Speaker
have some of the non-core things. It could be UX, it could be, ah know, a tech writer to document all the stuff, or QA engineer to validate the stuff.
00:59:50
Speaker
um ah You know, these these are all these are all some of the non-core things, right? I would rather have somebody that does ah better than me you know, because you can't do everything world class. it's ah it's almost It's almost impossible. And because you have a lot of competing priorities, that's where we

AI's Role in Quality Engineering

01:00:09
Speaker
come in.
01:00:09
Speaker
So we we now have the expertise in terms in terms of the people skills, the tools, the latest things, um while the individual companies are trying to figure out how to how to write a new sophisticated way of underwriting a risk.
01:00:30
Speaker
and they're thinking about the which markets to go exploit and then what is the next technology that they should be using. And there some of these things, they actually want somebody that is thinking about it day in and day out to do that for them. that that's how That's where we come in, Akshay.
01:00:49
Speaker
Okay. So, to, like, essentially to really have world-class quality engineering, it is, better to bring in an organization which specializes in quality engineering because a quality engineer that they deploy ah in your pod ah has seen multiple projects in the past across different domains. And but so that person brings that to the table.
01:01:18
Speaker
Absolutely. Absolutely. Okay. um see but but but I feel like on paper, yes. But a lot of Indian IT services are run by people who are one or two years out of college.
01:01:33
Speaker
ah Is that really something which gets delivered? like like you know the And how does it get delivered? like like like If you're putting someone who's one or two years out of college as a quality engineer in a team, would that person really have that world-class, that ability to do world-class quality engineering?
01:01:55
Speaker
yeah so just like anything uh actually like the the way that we go about the building teams is through a is through a resource pyramid and i mean you know so typically let's say when we work the bank or an insurance company they have the development teams are typically about you know 200 people, 300 people, 500 people. It can be very large. I'm just talking about within one business unit. And and the typical test effort effort that is being spent is say that if if there is $100 that is allocated to a project and the yeah and typically, ah you know, 25, 30, 35% of that budget is allocated for ah software testing and quality.
01:02:46
Speaker
Wow, that's a massive chunk of the budget. yeah Yes. And then actually for that, you then then you go about putting, basically thinking about building teams.
01:02:59
Speaker
And then you have how people that are project managing the testing activities itself because the testing itself can be very massive. And there are also people that are architecting. We also have roles like a test architect, just like a system architect, there is also a test architect, somebody that is thinking about how to design test, organize test, execute test, and things like that.
01:03:24
Speaker
And then you have ah test leads that are managing teams, and they've seen that engineer then then they're junior engineers, right? The test leads are the ones that are thinking at a module level and making plans and senior engineers are the ones that are mostly senior engineers engineers are the ones that are actually developing these manual test cases are automated scripts and executing them reporting results and things like that there is a proper
01:03:54
Speaker
resource pyramid that is being built actually the ones that you are talking about are the ones that are at the the bottom of the pyramid, you dares they constitute majority of the team, but they themselves um cannot think end to end.
01:04:12
Speaker
They are supported by by all these mid-level senior teams that are making plans and and architecting things and and things like that.
01:04:24
Speaker
like Okay. So that is the edge of outsourcing. While yes, you would be getting someone who's a year or two out of college, but there is a layer and multiple layers above him, which are giving him mentorship, which are giving him a standard operating process to follow, which are giving him tools, which are giving him training, and so on and so forth to ensure world-class quality engineering.
01:04:51
Speaker
Absolutely. I mean, you know, these are the these are the ones that had seen the world and I'd say in multiple systems, ah how how how they failed. and And also this whole quality engineering is ever evolving because a as as things are as things are moving to AI, as new newer neural newer tools are coming and even the quality engineering and software testing also has to cope up with it, right? Cope up with it and and also, you know,
01:05:26
Speaker
do exceedingly well compared to the software development. So there is that all natural push, right? And you if you are a product development company or an IT team within the insurance company, your first priority is ah is making sure that you know, am I getting my designer as my developers, my architects, are they using these some of the latest and greatest? And then, you know, for some of the non-core things like ah like software testing and others,
01:05:53
Speaker
can i work with a company that is already thinking about these things and probably can do um ah better than what we can do internally because because you know no insurance company is uh at the top people their their engineers are not getting out from the bed and then you know thinking about how do i create a best um you know the world class q practices within the company they are thinking about their business because that is that is what their age is but actually we are thinking about latest and greatest QE tools processes technologies and that's that's what we are thinking and that thought process and expertise is what we are able to offer to a set of clients actually is so testing easier therefore it pays less or I think it is
01:06:48
Speaker
i think it is it is ah it is ah easier to get started ah but not to become an expert right it is ah ah ah i think ah it ah ah it looks easy from a distance and and also probably there are less entry barriers to to get into this ah to get into this profession but then entry barriers in the sense that A code that needs to have proficiency in the language.
01:07:19
Speaker
Somebody without coding knowledge also can get it. Okay. So you don't need to be proficient in some coding language to be a tester. Therefore, entry barrier is lower.
01:07:30
Speaker
Okay. and and Essentially, it's more about logic, common sense. and like those It's more of those skills, like general aptitude and code. Okay.
01:07:41
Speaker
Analytical skills. but Okay. Okay. Interesting. So... This sounds like the perfect role for AI to replace.
01:07:51
Speaker
Yeah, yes. it I mean, you know even now, even if you look at it, even the software coding, I think there are people that are announcing it right there that 30 to 40% of the code is written by machines and things like that. yeah But yes, there is ah the some of these activities that are typically done by the ah software test engineer, quality engineer.
01:08:15
Speaker
they are, the they can be done by some of these a AI much faster and also with the reasonable accuracy, I would say.
01:08:26
Speaker
And actually we were the first ones to see it. So we started seeing this almost about a year before, and then we developed something called as a company, Cumentis AI.
01:08:39
Speaker
Cumentis AI is basically a quality, engineering AI platform that actually has 16 different features and each feature and addresses a task of a ah that is done in the typical software testing lifecycle. This task would be reviewing user stories with the intention of finding defects. The task could be ah generating manual test cases from user stories. The task could be executing automated test.
01:09:16
Speaker
The task could be doing a defect analysis. The task could be doing a risk analysis, things like that. So we we actually have built a platform And I wouldn't necessarily say that it could replace an engineer that is out there and, know, think of it as a more like ah kubernetes yeah is a Yeah, it's like a co-pilot for a software test engineer and, you know, so some sort of an ah expert assistant ah sitting next to you and helping you do your tasks. ah
01:09:51
Speaker
um You know, right now, I would say more quickly, but the but less accurately. But accuracy is improving every day as as an as more and more you use it, I think the accuracy gets better.
01:10:04
Speaker
Okay, so essentially what maybe three test engineers were doing earlier, now two can do with the help of these AI tools. I would say right now we're at about probably 25 to 30% productivity improvement. Okay. From five, it's come down to four. Like what, five engineers are doing? Five to three.
01:10:26
Speaker
Okay. Okay. Okay. okay Interesting. and
01:10:33
Speaker
But I guess the need for hiring freshers as the biggest chunk would come down, right? Because now that you have AI to do a lot of grunt work, typically, freshers are hired for grunt work, right? the The work which is not so intellectual, where you and which now AI can do.
01:10:54
Speaker
And so one senior test engineer can achieve more with an AI agent than he could achieve with a team of two juniors under him.
01:11:06
Speaker
that is a That is right, Akshay. I think the the the activities ah typically as the activities that are that are less skill intensive and then more effort intensive are the ones that are typically done by the staff.
01:11:29
Speaker
by the junior engineers in in any team, right? And actually, yeah, sort of can do those activities really well.
01:11:39
Speaker
i mean, yes, it does put a lot of pressure on engineers to sort of move beyond doing this, you know, grunt work, as you said.
01:11:52
Speaker
it it definitely It definitely is putting pressure on them to to move up in the in in the skill levels that they have. Is this changing your hiring strategy?
01:12:05
Speaker
Are you hiring less freshers or like i would say that I would say that we it will not change the strategy per se for us. I think probably the amount of time that we need to invest ah in training them on our platform.
01:12:26
Speaker
And because now now the ah the tasks the initial tasks or the initial activities that are expected to perform in the project have moved up so maybe maybe a bit more training is ah is needed for them rather than actually changing the hiring strategy actually ah because the because of the platform that we had built because we ah built our own platforms of probably the the number of training days and this platform will be higher
01:12:57
Speaker
okay Okay, interesting. So you built this with a, like through an API integration with a large language model. ah Is that how you built it? How did you build it?
01:13:07
Speaker
Correct. um We we have built this, which is the platform that that but when we actually started off like any other thing, then actually we were we were eating our own dog food.
01:13:20
Speaker
I mean, we are trying it out in our you know projects. and that turned out quite well, useful, and we experimented with the friendly clients.
01:13:33
Speaker
we We did POCs and the feedback is ah phenomenal. So we are now we are now going all out in the production. Also parallely, we are also upgrading the system that we have built to the enterprise and propris level.
01:13:50
Speaker
And ah so that way, like we are able to sort of allocate ah clean separation of the projects from multiple different clients, teams, and the authorization security is fully built in.
01:14:05
Speaker
And also that making it LLM independent that actually the clients can choose on the LLMs that they could use at the back end and also a mechanism for them to feed their own documents and so that they can effectively implement the RAG action.
01:14:29
Speaker
Okay. Which LLM did you find to be the best for this? Yeah, right now the GPT seems to be good. Open AI, okay. for software and quality engineering.
01:14:42
Speaker
oh Yeah. Okay. I heard Gemini was like the most recent model of Gemini was really good at software yeah coding. Yes. These these are ah these are very fast changing even. Yeah. It's a fast moving. Rock is also getting better. Okay. Right. It's getting better as well, especially especially on the technology front. But then actually,
01:15:11
Speaker
When we started, ah it was all GPT. So that's what we started doing it. But then now we are making it making it llm uh independent uh because the way that we work these things also we we were we also want to offer this platform to our clients and then so that you know they could use uh because you know whatever that we built it has to be a combination of software testing the technologies that they are using and also the industry vertical domain that they are into right uh so so it is ah
01:15:49
Speaker
We still don't know ah which LLM will dominate which industry verticals yet. I think that will take some time. So all we can do is stay neutral and have this platform work as an LLM um neutral so that so that actually we can then pivot whichever way that we want to go in the future.
01:16:13
Speaker
You use the term RAG. What does that mean? it's It's basically that that actually when you ping an LLM with a question, it comes back to you with the with a response.
01:16:24
Speaker
And then actually that response can be augmented ah with your own knowledge based on your internal systems. So then you you combine both of them for a much more effective due response.
01:16:36
Speaker
bye ah one that the one that you are The response that is coming from ah and you you adding based on your internal knowledge, that that the whole thing is there it's basically you are augmenting the response that LLM gave you with your own internal ah knowledge that you have.
01:16:59
Speaker
Okay, okay, okay. And know which you don't want to expose to the others because that could be, you know, your differentiator and don't want to know about it.
01:17:11
Speaker
Okay. Okay. Okay. ah Why don't you just... So you are earning... ah Your unit of revenue is a man hour so far? Yes.
01:17:26
Speaker
It is ah is a tied to hours. Correct. So which is why then you don't want to sell this AI agent as a bundled service because how do you...
01:17:41
Speaker
man hours when it's an ai agent and therefore you're pricing this as a separate separately price currently we are not even pricing it separately yeah actually we we actually are giving it as a value add to our clients i think we are at a point uh where where actually we we just uh we we are giving it as a a no cost value add to our clients we would like to where this hit your uh billable hours because if you are quality engineers are more productive like you know three people doing the work of five people so that means revenue loss actually you gain lot more compared to what you lose uh uh because uh because actually ah the the project that we are working on we may lose but then we will gain several more projects we are able to do it much more efficiently compared to the
01:18:37
Speaker
other vendors that are there in the client organization are their own teams and gradually we will get more and more work not only we get more work we're able to do it more efficiently and i think the stickiness all also with the client will improve so i would say that i would say that organizations that has very large test teams that are doing manual functional testing activities Like I wouldn't name the biggie names, but you can guess who they are.
01:19:09
Speaker
And I think it poses a lot more risk to them than to us because ah you general home no we probably we probably don't even have 1% of the entire testing market. Right now, we probably are at about,
01:19:25
Speaker
yeah 0.75% or 0.5% of the entire testing marketplace with us. So we, mean, you know, either we can, we can look at this 0.5 and then see what do we lose, or we can look at 99.5 and then see how much is there to gain.

Price Dynamics in IT Services

01:19:40
Speaker
Fascinating. I do remember reading somewhere that in these kinds of situations where there is massive change of technology, it's the not the downstream users of the technology, but the technology companies themselves who capture most of the value. Because the downstream users kind of, you know, that one would think that you would have more margin because AI is making you more productive, so you need less people. But then the competitive pressures kind of but are at play and therefore
01:20:15
Speaker
the the most value is captured by the the, like the arms dealers are the ones who are capturing the value in an arms race, basically. Correct. Correct. Yeah. and And, and, and also, and also saw that if, if we don't do it, somebody else.
01:20:34
Speaker
Right. That, that, copy that competitive dynamic prevents you from increasing your margin as a result of More efficiency. Got it. Interesting. Very interesting.
01:20:46
Speaker
and And it also makes it so much easier to go and ask for a price increase as well. Especially if you improved efficiency, that is a much easier discussion compared to that I had spent two years and then I didn't get a rise.
01:21:04
Speaker
How do... price escalations normally happen are they baked into a contract like when you sign a contract do you get X dollars per man hour for year one and then X into 1.1 dollars in year two or like how does it happen?
01:21:21
Speaker
they The price increases are sometimes are part of the contract and sometimes they are not and you even even if they are part of the contract we still need to ah take concurrence from the client and and make sure that they're uh they're okay with it um that's that's number one because i would say that uh much of the uh ah I mean, I think though the way that at least we manage is is basically becoming more and more of efficient efficient from year over a year as opposed to um as opposed to getting these rate increases.
01:22:02
Speaker
Because if you if you do these rate increases like... ah like a consequent consecutively for ah three four years you'll essentially get priced out of the market that's a serious risk that you will run into so i think the best strategy is to basically ah how could we offer a higher value for the same price as much as possible i think price increase of you know two percent three percent are pretty common in the industry but if you try to do anything more than that if you do that oh say four or five times
01:22:36
Speaker
you know, you'll get priced out.
01:22:39
Speaker
Okay. Okay. It sounds like really cutthroat industry. ah Like a 2% price increase each year. A lot of competitors are eyeing the business that you are getting from the client.
01:22:57
Speaker
Yeah, it is, ah it is it is ah ah i mean, and also depending on the clients that you are working with, there are some clients that are price bias and there are clients that are value bias, right?
01:23:09
Speaker
And you will very soon find out, i mean, you know, once you step into the relationship, start working with them, it's a matter of for two, three years that do you know who is whom because everybody tells you that they're all value bias.
01:23:21
Speaker
flash yeah Right? And... the depending on whom you are dealing with, I guess we really need to adjust ourselves.
01:23:33
Speaker
and so And so, like, you I think it's easy to, I would say if you want to build a long-term sustainable organization, it's better to work with value buyers because they are constantly ah pushing you to do better.
01:23:55
Speaker
Whereas the price price buyers are consistently pushing you to to lower the the quality of work that we do, and which is not ah which is not really interesting after a period of time.
01:24:12
Speaker
Okay. Interesting. so so like that the the yeah Demanding customers who make your life miserable are the customers that you should pay the time. Yes, yes, yes.
01:24:22
Speaker
The ones that are that are really that are really pushing it, like the one that you talked about, right? Like, yes, I mean, yeah, all that is great. But then if I have a 10-member team, can you do this the same amount of work? Assume that the work is not expanding. Can you do it with eight people?
01:24:37
Speaker
I think that's ah that's a great discussion because then that keeps you thinking about, optimization and tools technologies and everything as opposed to somebody that just asking you to reduce the the price and and you know they have the any option that you take to reduce the price is not helping in the long run but okay for them and also for us for the client and also for the for the service product
01:25:08
Speaker
okay okay Okay, interesting. um Okay, so one aspect of AI is the AI replacing or making quality engineers more productive. The other aspect of AI is how does the AI change the fundamental nature of quality? Do you also do quality engineering on AI itself?
01:25:35
Speaker
Or like what are the implications of that for your quality engineers? Yeah, but that's a great question. we We recently won an engagement with an insurance company where actually um they actually have built systems on the latest LLMs and chartboards.
01:25:52
Speaker
Now we are being involved to test those systems, which is essentially testing systems You're testing for hallucination. and Like the big problem with AI is hallucination. Like it gives wrong information.
01:26:05
Speaker
If it can't find an answer, then it will bullshit its way to an answer. ah so So you are testing for that? we Yeah, so yes, we are testing for that. We we we are now we now started testing systems that are that are built using LLM systems, where the stuff that you talked about, hallucination, and also that, i mean, you know for example, the systems that were built so far without AI were like, you know, um you feed in 10 different values into a system, and then actually there is only one right output.
01:26:44
Speaker
right vita and the the the the and then you get that it's passed you don't get it it's a fail but now the AI systems don't work like that they work on a range of outcomes and then also accuracy levels and all of that so it it it poses a different challenges actually and is there ah risk of a product replacing you but you are essentially a services business and you charge per man hour or per man day or so something to that effect.
01:27:22
Speaker
um You could have a pure product company come in and eat your lunch. Do you see that happening?
01:27:30
Speaker
o I mean, there is, ah I would say that right now, right now also as part of the service, we many different tools.
01:27:41
Speaker
Not everything, know, that is being done by us is all isal manual. So there are there are a lot of different tools that are being used.
01:27:52
Speaker
if And I think the role of platform as a tool in the quality engineering space going forward um oh will be continue to go up.
01:28:07
Speaker
However, ah there are some tasks like planning strategy and thinking and reporting oh with some but some added insights, things like that will will always be there, Akshay.
01:28:30
Speaker
oh but But yes, I mean, you know there it will always be there is always that of tools, a risk call for ah tools ah sort of start taking over some of your activities. But that but I don't see a scenario um where actually yeah the tool replacing all the activities that are performed by the quality engineers.
01:28:59
Speaker
I mean, you can never say never, but the the possibilities are extremely rare. OK. um Possibly the SMB market could be served through a tool, I guess. Maybe not the enterprise market. And this is typically how tools evolve, right? Like, say, Canva as a design tool or Figma as a design tool all had SMB adoption before enterprise adoption. And my other counter here to you is that ah could you be constrained by your
01:29:33
Speaker
ah experience and your, I mean, you have a lot at stake in terms of the fact that ah humans cannot be replaced. I mean, your business runs on that, right? Could that be biasing your take, your view that products will not replace humans?
01:29:50
Speaker
Yeah. So, yeah, that could be there. But ah however, even even internally, we the way that we we have started differentiating ourselves um since the inception of the company is to build uh tools utilities platforms accelerators to continue to disrupt ourselves so there had been the motto um right so we we are always challenging ourselves and uh and building
01:30:26
Speaker
a tool or utility to do what we're doing. so ah so So internally also we are yeah had we are pushing it as ah more and more clients are demanding for us to become ah more efficient and more innovative.
01:30:42
Speaker
So we we certainly cannot escape from that. And also that for the kind of work that we do, we also stay in touch with the clients almost on a daily basis and and some of the decisions around their investments in tools and all that majority of the time they are joined but we also have we also have clients that that they go and ah do on their own as well so yes i mean you know even you
01:31:18
Speaker
you If we resist ah ah some of these genuinely um ah possible, ah genuine improvements because of the the tools getting involved, i think ah I think our competition will pick it up.
01:31:39
Speaker
ah we We would rather disrupt ourselves rather than allow our competition actually disrupt us.

Efficiency Tools in IT Services

01:31:47
Speaker
That's the mode of how we think about things actually. Okay, understood.
01:31:52
Speaker
um What is an accelerator? you know I had a short stint in a services business. I do believe for services businesses, this is one of the ways to differentiate themselves by having accelerators. Can you help me understand what is an a accelerator and how do services and particularly IT services companies use this as a differentiator?
01:32:14
Speaker
Yeah, sure. surera So as i and SAP is a is a fairly popular and one of the most a used product that is out there.
01:32:27
Speaker
And you know it is used to basically to run... multiple different processes within the enterprise. and And for example, a large enterprise has has implemented SAP and they have automated their entire say like their purchase and payment processes to be run through sap so so every a organization is different on their own but but not so different ah as well so what what we would see is that typically about
01:33:07
Speaker
80% of the processes are common, even them among even within different industries as well. So the accelerator is something that we have a a pre-built ah manual test cases and also automated test scripts to validate that to validate these processes so when when actually somebody a large enterprise client is approaching us ah to to test um the the the processes for which we already have the pre-built accelerators in a sense uh 80 percent of the uh the work
01:33:47
Speaker
ah could be done with already the pre-built and ready to go kind of test assets are already available with us and and there is uh there is always customization that every every organization is different to to that extent only uh we we really need to customize and and change things so i would say having an a accelerator it gives you that jump start uh that is needed yet because you don't have to build it from the ground up and another another thing is that uh
01:34:24
Speaker
you use accelerator multiple times in different contexts, it gets better. So you are actually using something um ah that is actually used many times and sort of tried it with the many different clients. So the possibility of it actually working out um is is really, really high.
01:34:45
Speaker
Okay. And does this also save man hours for the client, their billable hours? Yes. Yes. It ah deep could save, I would say, as much as 40 to 50% of the overall testing hours that need to be spent in a large program.
01:35:04
Speaker
Okay. But you would offer this free of cost because it's more important for you to win the client rather than monitor. Yes. like necessary That is absolutely right. I mean, okay yeah. Okay. ah Okay, okay, okay. Very interesting.
01:35:18
Speaker
And you have like an internal team which is maintaining these accelerators, improving them each time they're used, or is it the operations team only which does it? So we, you know, the operations teams, they can get involved, but their amount of involvement could be limited because because their time is ah spent on the client projects.
01:35:40
Speaker
They don't have the kind of bandwidth that is needed to build it. So we have something called Centers of Excellence, ah and which is basically an in-house team and that is ah staffed with a dedicated resource.
01:35:57
Speaker
And they are the ones oh that are working on these things, actually. And they typically, the Centers of Excellence that that we have in the company are, we have oh you know, particular service type.
01:36:14
Speaker
Like we have a centers of excellence for performance testing, for security testing. And we also have oh so centers of excellence for domains as well, ah for any industry vertical like insurance, retail, travel.
01:36:33
Speaker
And we also have centers of excellence based on the ah technologies, like Like, you know, we have centers of excellence for DevOps, for AI and things like that. So centers of excellence is really three different three different angles that and that we take Okay, interesting.
01:36:57
Speaker
ah In your wage bill, what is the build to unbuild ratio? So you would have like build resources which are directly being chargeable man hours being generated by them and then you have these centers of excellences which are like unbuild resources, right? So what's typically the ratio? like ah So if, I mean, we are not very different from every other IT services company that is out there actually. So, but of course, since we are still young and growing at a very high pace um our utilization ratios are i would say probably by five percentage points lower uh compared to um you know the companies that are already big in size like 10 000 people 15 000 but what is that number but i'm not available will be about 75 percent okay
01:37:47
Speaker
okay So 25% of the headcount is in centers of excellence, building accelerators, or sales, marketing, these kind of roles. Yeah, I would say that also that many of them, some of them are also being and working as buffers in the projects to actually to provide coverage people.
01:38:11
Speaker
um that are that are going on vacations. And also there would be there would be times oh where actually we also need this kind of a capacity to ah to deal with the some unanticipated events like risks in the project.
01:38:32
Speaker
And we also need, ah especially, we also need a certain amount of for people um that are waiting to be deployed no with all kinds of designations, right? and Like ah a representative of the teams that we typically deploy to our clients.
01:38:51
Speaker
So basically it's are some of the bench also to to to really to get going with the client projects sooner than later.
01:39:04
Speaker
Okay. Okay. And this you're saying in... Like you said, some R&D activities and COEs and they're also part of this. Okay. Okay. Okay. And this you said in large organizations is 80% currently for you. was a d Yeah.
01:39:19
Speaker
okay It's anywhere between 80 to 85. In our nation like us, it is more of 70 to 75. Okay. Okay. okay And ah how much ah is the monthly billable amount for one headcount?
01:39:33
Speaker
that you charge? Yeah, again, it it' sort of varies based on each role. And we actually, i mean, as a company, um it's anywhere between the the most or junior person, which is basically the software test engineer in a team and all the way up to ah like a program manager, right? And anybody in between like a test lead and test architects ah and architect and things like that.
01:40:03
Speaker
so So at ah at an offshore, um it could be anywhere between $25 all the way up to $45.50, right?
01:40:14
Speaker
This per man hour? Per man hour. At onsite, this could be, it could start from... say say 75 80 and it could go all the way up 140 150 right so i would say that as a company we are we are right in the middle i mean you know we are we are not We are not the the lowest price providers, but at the same time, we are not we are not cheap charging um the most premium provider as well. I would i would say that we are right in between and and actually where we like to think that we are sort of, that's where you can provide the highest ah value for the money.
01:41:05
Speaker
so what You know, what percentage of your revenue is the wage bill? Is that the biggest head in like? Yeah, so several of the engagements that we do right now, they're all based on TNF, time and material mode.
01:41:22
Speaker
That's right. And that, I mean, we we do have a few milestone based and fixed bid kind of projects, but they're relatively...
01:41:34
Speaker
both in terms of the effort and also the revenue contribution is very less, would say single digits. ah So majority of the revenue is TNM and that is also because of the way that we want to work with our clients. We would like ah clients to work with us on a long term basis as their extended long-term strategic quality engineering partner.
01:41:58
Speaker
so So, which means that, you know, just like the way you you work with your internal team, as the priorities change and as the direction changes as a company, and then actually, then you are able to move them to other projects and quite quickly.
01:42:16
Speaker
So it is the same model and that we follow. so i would say in majority of the times, the way that we work is that we are working with organizations as their extended team. So the projects that we work keeps ah changing. I think in from that prospect perspective, I think T&M makes the most sense.
01:42:42
Speaker
So I want to understand the margin profile of this business. So say you, hypothetically, if you do $50 million dollars of revenue in the current financial year, ah can you break it down? How much of but how much would be your wage bill for a $50 million dollars revenue? How much would be other expenses? What kind of margin would there be?
01:43:03
Speaker
So in anti-services companies, ah you know there are no real they not real expenses um outside of ah sort of salaries.
01:43:15
Speaker
right And so i would say majority of the expense is he's basically the wages for the consultants and employees, or basically call them as associates.
01:43:28
Speaker
right and What is the percentage number? like Yeah, I would say the typical gross margin um of the IT services companies is anywhere between 30 to 40.
01:43:43
Speaker
thirty two ah forty and there are companies that are doing above 40 but then i i guess you'd see very small amount of companies that are hitting 44 45 so that is the that's the grass margin i would say in terms of the the total wage bill for for all the associates and consultants i think it it would probably about uh 70 of the the revenue um would be in terms of the wage bill
01:44:18
Speaker
You said 40% is the margin, so that means 60% is left. Of the 60%, you're saying 70% is the wage bill. on Yeah, the what I'm saying is that if you take the total... I mean, not everybody is ah is billing and getting the revenue for you, right? And we also have people that are in non-billing roles.
01:44:40
Speaker
But if you typically if you typically look at the P&L of IT services company, you have revenue coming in and there are people that are directly uh contributing uh to the to the revenue by by actually spending all their hours working for clients and so if you if you look at those those hours those costs then the gross margin is typically about 40 percent and and then actually there are then
01:45:13
Speaker
you know If you go down the P&L, then you would actually have people that ah that are actually supporting and the individuals that are working in these projects in terms of ah the management support and tech leadership support and and things like that. Then you also have people that are not billing like we talked earlier.
01:45:38
Speaker
um you know the the over the the pinch uh bench count and uh and that's that's one layer right which probably is about up to five percentage points and then you have and then you have the biggest uh the head is going to be the sales and uh sales marketing pre-sales all these costs which is typically called as s and m and the the s and m is uh um is also in the industry varies based on your growth. I mean, if you are if you are growing at about 25, 30%, then the S&M needs to be as much as 10, 11, 12, it can go as has 14% as well.
01:46:24
Speaker
And then you have G&A. And the last time that I and took a look at it of the typical IT services companies, it's probably about yeah seven and seven eight nine ten uh essentially um you know you are a lot of it is really your uh um your ah your rental costs are um the the that kind of physical infrastructure that is needed for the people to work and then of course the the support function cost and you know that's how the typical p and l of 90 services company
01:47:06
Speaker
So typically EBITDA would be about 20% or so, I guess. Yeah, it it depends. I would say that what is did it's a large it's a large spectrum ah but based on and also depends on the the type of service that you are offering and also that ah rate which you are growing, the revenue growth.
01:47:32
Speaker
and also the quality of the revenue and also how optimized your internal operations are, especially in the context of economies of scale.

International Market Dynamics

01:47:41
Speaker
would say it is anywhere between 12, all the way up to 25%.
01:47:49
Speaker
But then yeah that's the that's the that's the range you are dealing with, actually. Okay, okay, okay, understood. Okay, okay, understood. ah Do you only sell to US companies or do you also sell to Indian companies? Is there a market for these services in India?
01:48:06
Speaker
Yeah, there is ah there is a market for these companies in India as well. However, the the
01:48:18
Speaker
India is a tough market because you you get your revenue in rupees and you spend your money in rupees.
01:48:29
Speaker
And so's ah it's a the margins are fairly low. And also even your Indian clients also have access to the same talent pool that that you also have access to. so it's So the the the margins are ah pretty low. Of course, there is there is an opportunity to do business in India.
01:48:57
Speaker
and Right now, we do have a few a few clients in India, but but we mostly do the business to build ah capabilities in certain areas and also to have some solid case studies so that we can showcase it to US clients.
01:49:21
Speaker
Okay. okay okay Got it. um Okay. I want to end our conversation with a little philosophical discussion on quality. um what is the How do you make out if a product, and particularly we can talk of software product, is built with quality or lacks quality?
01:49:44
Speaker
like like Is there a way to look at a product and form an opinion about the level of quality thinking that went behind the product? Yeah.
01:49:56
Speaker
yeah I would say that many times the quality is that you would actually perceive the lack of quality more quickly than the existence of quality like with like with anything.
01:50:12
Speaker
um So ah that is one. And also that no what we had what we had seen as the industry right now is that um a lot of these quality engineering teams they are playing the role of an uh sort of an upstream co-maker uh meaning that lot of testing activities are shifting left they are working with uh
01:50:45
Speaker
development teams hand in hand to build the quality into the product as opposed to yeah improving it by by testing it out later on right and so which is essentially the mentality of a downstream subcontractor So if you if you if you really look at by philosophy wise, so do you want to be ah downstream subcontractor now where the software is being ah thrown at you over the wall and then actually you you are now challenged with the
01:51:26
Speaker
finding everything that you could and then throw it back. That's one way of doing it. Another way is actually um just doing it together um even before a single line of code is being written. So I think... ah There was some report, I just can't recollect the whole thing, but then, you know, just talk about the the amount of you know money that was lost in the economy by not testing the software products um you know adequately.
01:52:02
Speaker
So it was a huge amount. It's like 500 billions or so. like five hundred billions so so so so this is so like a global number or for India it's a global number but okay and it was a global number so so what it what it what it still means is that uh the ROI of testing it early and testing it often and also testing it thoroughly is humongous and that is the reason even if you look at it if you read some of the reports it clearly shows that ah the amount of
01:52:41
Speaker
the the rate at which the the quality engineering the spend on quality engineering is growing is 2x compared to the the investment that is happening on um software development so you know my sense is that it is still under invested everybody knows the importance of it ah but however And they are doing, i mean, every other company recognizes this fact and they are investing.
01:53:17
Speaker
Even after this investment, it is still not

Creating Organizations Rooted in Quality

01:53:21
Speaker
enough. and And that is the reason why why actually I think it will continue to grow at a much higher pace compared to the software development.
01:53:32
Speaker
Okay, interesting. How do you build an organization which has quality in its DNA? Like, you know, where everybody cares about quality. Yeah, that's a very, it's a very, it's a very difficult thing. And think it's, it's more of, I think it's more of a mindset.
01:53:52
Speaker
And I, I guess the, the companies that are more client centric, ah I think ah they do a good job of building a ah the dna of the company to focus on quality then actually you are seeking something else i think ah you know you can have all sorts of philosophies and metrologies but what i had seen it working is that if you're if you're client-centric and i guess ah naturally you tend to do
01:54:35
Speaker
um high quality work and actually you take your deliverables lot more seriously you take your interactions lot more seriously you plan for them you strategize for them and and i think that i think that's the best way to ah build that the the quality as part of the ah dna of the organization actually are indian companies more quality conscious or chinese companies
01:55:05
Speaker
i I haven't actually the the industry that I'm in, in IT services, they're not roll off a lot of Chinese companies um that are operating

Global Perceptions and Industry Focus

01:55:20
Speaker
in this space. I would say that majority of the majority of the um ah the quality engineering in IT, the companies are oh they're from Europe and India and US, I haven't seen
01:55:41
Speaker
like the companies that are of a decent scale coming from China. I mean, China is obviously a manufacturing exporter, whereas India is a services exporter.
01:55:53
Speaker
That fundamental difference is there. But generally speaking, in terms of the organization DNA, you know, do you think that obsession for quality as an organization, DNA is more in Indian companies or more in ah Chinese companies. And I mean, you know, feel free to tell me that this is that you're not the right person to ask this if you think that is the case, but I'm just generally... yeah yeah i would say if i had to if i had to give my opinion that even a lot of a lot of processes and practices ah that are used in that are used in software quality are essentially borrowed from a manufacturing industry yes ah so so the i mean you know a lot of these ah the
01:56:44
Speaker
japanese manufacturing quality techniques are actually the ones that are used in even in the software and software quality so i would i would say that even even internally we say that we have this ah slogan that says trust the process results will come right and so i would say that if I mean, they definitely have that edge, and right? if If companies from China ah has to make an entry into quality, I think their pedigree of building something in the manufacturing space ah will definitely ah give them an edge for sure.
01:57:31
Speaker
Okay. As a country, do you think India is has a DNA of quality? ah Yes, i i think I think that... I mean, generally the perception in the world is that indian products are of inferior quality, right?
01:57:50
Speaker
So that's why people want to buy imported stuff, etc., etc.
01:57:57
Speaker
but I think when it when it comes to so when it comes to software and assuring quality of these products, ah the the products or the applications so that is built by the enterprises.
01:58:11
Speaker
i I definitely see the work ethic and in general, wanting to do a um better job is a definitely increasing actuallycha in the IT services space.
01:58:28
Speaker
um But I guess in terms of the the the tangible real world products that we use ah day and day to day, um I guess there is still a scope to improve.
01:58:40
Speaker
Okay. Okay.

Entrepreneurial Advice

01:58:41
Speaker
Understood. ah Let me end by asking you for some advice for future entrepreneurs, somebody who today in 2025 is planning to start a business.
01:58:52
Speaker
Do you have any advice, recommendations in terms of what are the opportunities to look at what should be things to keep in mind when you're starting a business, you know, anything at all from your end.
01:59:05
Speaker
Yeah. I would say that, uh, somebody, that is wanting to, ah ah start something, always start with, uh, I would say that, uh, uh, an idea, uh, that, that you are very passionate about, um, because, uh,
01:59:25
Speaker
because not everything may go the way that you you want it to go. And and or during those times, I think if you're deeply passionate about something, I think you will give it a lot more time um for it to work.
01:59:44
Speaker
And ah that is definitely one. And ah number two is that I think have to be really honest to yourself about you know, i are we doing this better than some of the existing players, right? Especially always always think about the value for money.
02:00:06
Speaker
Could we provide more value compared to the other existing players? And really hone in on the ah the differentiation aspect is is important.
02:00:18
Speaker
and And I would say third thing is that, ah you know, basically that you need to have paying clients.
02:00:28
Speaker
And so so identifying them very early on, even even without you know identifying those well-wisher and friendly um kind of a category of the clients, at least two, three of them before you you jumping all into it.
02:00:54
Speaker
ah would definitely help you a lot. I mean, you know, these are these are some of the things, ah you know, people have to think about, ah generally speaking.
02:01:07
Speaker
i mean, you know, do we need to check all the boxes? ah May not be, but the more boxes I think that you check, I think the better chances success, Akshay. Okay.
02:01:18
Speaker
Thank you so much for your time, Kalyan. It was a real pleasure. Okay. Thank you, Akshay. Thank you very