Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
How Generative Scheduling is Accelerating Construction Planning
82 Plays3 months ago
This week, Jason & Carlos speak with Georgia Stillwell, Director of Client Solutions at Alice Technologies, an A.I. driven construction scheduling tool.
The trio delve into the unique approach Alice takes to A.I. scheduling, the difficulties of existing legacy tools, and how much further A.I. can be applied to construction scheduling.
Follow Carlos on Linkedin | Follow Jason on Linkedin | Check out Aphex
Recommended
Transcript
Introduction and Iterative Scenario Planning
00:00:00
Speaker
So it's quite iterative. It's unlikely that you run one scenario on Ellis and it's you've nailed it. You've got the perfect project. It's you go check it. oh Actually, this is where the kind of critical work is. Let's focus on parametricizing just that section and see where we get to. So it's a exercising that's a summit. That's 100% the biggest word ever said on this podcast. ever
00:00:35
Speaker
Welcome back to the offsite podcast. Carlos, how are you today, mate? Pretty good. Thanks. How are you? I'm very good. I had a peek at your calendar earlier, uh, cause I was looking for a slot to put a meeting in and I noticed that there was a block out the other night for football. And for anyone that knows you, you've got a sorted history with football, which is like, as you go back to play and then you damage a knee, shoulder, hip, something. Um, are you back playing football in the evenings?
00:01:03
Speaker
Yeah, the last time I played it was nearly a year ago where I tore the ah socket off my shoulder. Full surgery, long-term recovery. ah Yeah, no, I played last night for the first time. My knees hurt and I feel old. I can't remember it.
00:01:18
Speaker
Didn't think about like another hobby given the track record? No, well, yeah, I enjoy it. I do need to stop. My body tells me to every time. And then I have to use the, I have to get on the peloton to recover my knees just to play the next game. So I said, I should just stop. Do you just get stuck in keeper or are you out and about?
00:01:36
Speaker
No, I can't do keeper. I haven't got the, I haven't got the replay. It's not even sport if you keep it really. Like I don't want to be like slating keepers, but yeah. Yeah. Yeah.
Curiosity about New Tower in Saudi Arabia
00:01:46
Speaker
Um, also, did you see in the news? Well, I don't think it's that recent, but ah it's only really surfaced for me, but have you seen the new tower of the building in Saudi Arabia?
00:01:55
Speaker
Is this the one that's like 400 kilometers high or something? It's like literally trying to... Yeah, it's the line that just decided to go vertical. It's 1,400 meters high. right that's That's almost five of the shard on top of each other. And it's like 400 meters. It's over a shard higher than the Burj Khalifa in Dubai, right?
00:02:16
Speaker
Yeah, wow that's nearly a mile. but So yeah, will it be built? We'll only know for sure if ah if Tom Cruise jumps off of it and like Mission Impossible 15 or whatever it is by then. um I don't know what to believe with the yeah project announcements. Right, let's jump into it because I don't want to say anything controversial.
AI-Driven Scheduling Tools Overview
00:02:34
Speaker
um So today we're going to dive into one of our old favorite topics. Well, definitely one of my favorite topics, which is scheduling. In a bunch of ah the circles of people that we would talk to on a daily basis, Carlos, I think AI and scheduling is probably best described as like pineapple on pizza. Some people love it and some people are mega skeptical.
00:02:58
Speaker
However, I guess in the cohort of people that are probably in the more skeptical camp, there's one AI-driven scheduling-based tool that just always seems to have positive feedback from anyone that ah interacts with it. And so I would like to talk about that today, which is Alice Technologies. And we have the, I guess, perfect guest, Georgia Stillwell, um who's the Director of Client Solutions at Alice. Georgia, thank you very much for joining us today.
00:03:28
Speaker
No worries, thanks for having me. So I think we'll dive into, obviously we'll try and dive into what Alice does, how it works. But before we do that, to not ask the sort of standard question, I'm sure you get asked 15 times a day, I'd probably like to zoom out a bit and talk about how do you or how does Alice view the future of ah schedule management on projects to maybe set that conversation up a little bit for you. People kind of sit in two camps that we would interact with on ah on a daily basis, those that have a strong opinion. There's this cohort of people that think like the future is very top down, that the way that the project will operate in the some distant future is that there's a schedule that's like being constantly optimized with
00:04:16
Speaker
AI or other methods, and that that's really directly connected to what's going to happen on site, driving resources, you know allocating resources to things. And that kind of the need for the dirty on the site coordination piece gets less and less over time. And then there's this other camp which says there's no way that all that detail can be known in the schedule. There's always going to be some logistics and some stuff that's never captured in the schedule.
00:04:42
Speaker
really, the future is better tools for the people on site that need to deal with that unknown unknowns problem. Do you have a do you have a view about the future of that? Yeah, so maybe this is a politically correct answer. But I think both are needed. um I should have said that. We want to hear the non politically correct answer to this Because I believe you know with computing power we can capture so much detail and crunch it and work through it to find you know the best way to optimize you know an excavator on site and teams that are on site to keep
00:05:21
Speaker
productivity up, time on tools up, but there will always be nitty gritty details on site that that's why we have site managers and sub agents and these people there because there are little bits and pieces that will never be captured by a system or a model. I think it's both hand in hand and ideally they both elevate.
00:05:43
Speaker
are elevated with tech.
Deep Dive into Alice Technologies
00:05:45
Speaker
And so it's not kind of writing pieces of notes on paper and like doing verbal handovers to whoever's coming on site next, but actually capturing it in an intelligent way and making their job easier to manage those little bits and pieces ah that come up.
00:06:01
Speaker
How far do you reckon it can be taken? I guess the um the the schedule has the tasks that need to be done built into it. There's some there's some physical model. um There's some resources in there. How much further does it need to be taken or can it be taken in terms of understanding more of the context of what actually is happening on site?
00:06:20
Speaker
Yeah, so I think it can be taken pretty far. Actually, it's more our imagination or perception of tools that probably limit what we can do. The thing is, if you want to add detail, you've either got to collect it some way and and unless that selection is automated, that then becomes an odorous job for a human.
00:06:37
Speaker
and And is making that data meaningful? So often ah people think, oh, let's just add all the data to Alice and because we can and we can crunch it. But if there's a subcontract and it's all been in grade, do we really want to go and meddle with maybe what their methodology and plan is? So why don't we like leave that part, almost freeze that part and focus on where you know, there are gritty interfaces or complex work where they do want to detail and we're having AI and and a machine kind of crunch and discover options will be helpful. So I think being targeted with where you use data and like where you would actually focus optimization rather than like, let's just optimize everything all the time ah is probably a better approach and more andef efficient efficient as well.
00:07:25
Speaker
Yeah, I think that's a really great and and practical answer. To maybe then zoom back to the question I probably should have asked at the start. For those that don't know of Alice, I'd say a lot of people that that I've interacted with have heard of it or seen demos or used it on projects and are very interested. But for those that haven't, do you want to kind of give the one ah one at the high level? but yeah Yeah, that would be awesome. and And definitely not for me or Carlos, because we know it in detail. so Yeah, so Alice, it's ah it's a software tool with AI for scheduling. So we say generative scheduling or parametric scheduling. So by that, ah you're giving it rules. So construction rules, right? Like we've got to excavate before we do the foundations. That's a rule that you have to tell Alice. And then anything that isn't you know a hard construction rule, so whether we start north to south, east to west,
00:08:24
Speaker
work in different areas, how many people, how much equipment, those things that are flexible, the AI will then explore those options. So literally run like a Rubik's Cube, all the different combinations of how your scope could be put together with those different variables, whether that's resources, equipment, material, actual sequencing, and coming up with many different schedules or plans. And so from that, you can pick whatever schedule is ideal for you.
00:08:53
Speaker
is what we find today is with legacy tools, which I'm sure many of your listeners are familiar with, they're often so difficult to use that people only put together one plan. And then, especially when you go on site and to build, that plan is so rigid and there's like a million links and it's like spaghetti that, you know, you can't manage it. People go to other tools or they go to Excel.
00:09:16
Speaker
And they do it another way because it's not agile. So with Alice, we're making many plans and also this kind of agility. So if something changes, you can then re-optimize from that new point in time. A little bit like a GPS. So you make a wrong turn. The GPS doesn't tell you to reverse. It then checks all the ways to get to your end destination and then gives you the fastest one. So that's what Alice is doing as well. you There's a delay on site, maybe a late material delivery.
00:09:44
Speaker
And now we're looking at with all those variables that were flexible in your project, which one should we dial up or down to still meet that end date? I think that ah will resonate with almost everyone listening. I think like think there's so many projects I've worked on in the past where you get to a certain point in the project and you're thinking, well, what if we did this? What if we changed the access sequence into this building? What if we did this before this thing?
00:10:07
Speaker
And then someone goes and asks a planner, and some of them are quicker than others. Probably the polite way to say it. And then some days or weeks later, you get an answer back on one of those. You go, well, that's not it. What if we did this or this? And then you end up asking an engineer to go and maybe do it in a, I don't know, a quick Microsoft Project high-level version of it or something like that. So that definitely resonates.
00:10:32
Speaker
Or if I break it down into components and do kind of like a back of a fag packet assessment of if the scope would be here. But again, it's quite imprecise. It's difficult. And is it robust enough?
00:10:48
Speaker
to actually execute on. We've seen a bunch of our customers trying to do that using staging diagrams, kind of marking up you know a slide deck using that. And they're like, we we're trying to plan for the next three and a half years ah in in your short-term planning tool. but Are you, what are you doing? um So yeah, now we know who to refer them to. mas oh no And then he is like a PowerPoint or something. And then like the excavator appears.
00:11:15
Speaker
So in terms of actually, if we're using it on site, so you mentioned how we sort of we plug in the schedule, we apply some rules that could be like excavation before foundations. And then we're looking at various scenarios that that come out the back end, which we then choose which one we'd like to potentially implement. Do you constantly sort of refine those rules? Because there could be limitations on things like that we just can't have this many people or we can't switch to this type of plant. So every time you run the schedule, you add more rules and eventually you get to one within the the sort of constraints that you have to work with. Is that how it works in practice?
00:11:52
Speaker
Yeah. So there's actually two ways you can come about it. So you can do the kind of carte blanche or blank canvas approach where you literally bring in the scope. Is that your and friend? That was your French Georgia. Yeah. Yeah. fli got Where you put in like the minimum rules, really just the construction rules. And it's very free and you can explore so many options.
00:12:16
Speaker
Often you only have that flexibility in like super, super early stages, like before you're on site, like in a tender where you were coming up with the strategy. So you can come about it that way, carte blanche or blank canvas, and then add rules. And it's taking like a few clicks to add rules and like two minutes to run scenarios, right? So it's not so difficult. Uh, it's not onerous at all to then add in rules bit by bit. The other way you can come at it is we have a direct import with most legacy planning tools. So you can bring in a plan.
00:12:47
Speaker
which has been highly constrained, every link is hard, and what you can do is kind of renovate it. so We have like bulk tools to remove preferential or soft logic, and you can kind of bit-by-bit scalpel approach, create flexibility in an existing plan, and that way you might explore just a few elements as opposed to exploring the whole solution space with the kind of blank canvas approach.
00:13:12
Speaker
So you can come at it either way. I liked the constraints you mentioned, which was like density or space. So that's something we can easily model in Alice. So it's quite a cool approach.
Focus on Project Objectives with Alice
00:13:22
Speaker
You say like in this area, we can only fit 80 operatives.
00:13:27
Speaker
but run an Alice, like what combination in which crews should be there without prescribing, like there's going to be six electricians or like four, and it will find the ideal mix within that constraint of 80. So that's a good example of implementing a rule, but then still allowing a flexibility of what mix and like what activities should be happening in that time.
00:13:47
Speaker
And can a rule be like your objective might not be to shorten an end date. It could be to reduce carbon or reduce cost or something like that. Is that also sort of coming to the mix? Yeah. So anything that's numeric, we can like point the algorithm towards. So by default, it goes for duration and cost, but if you, you can also reorder.
00:14:07
Speaker
what you want as the top kind of objective for the algorithm to focus on. So we have some clients in the UK are working on big solar projects, and their huge concern for them is carbon emissions and peak carbon emissions. So what they can do, the same way that we talked about creating a constraint for 80 people on site, we can create a constraint for maximum carbon emissions at any point in time and use that to create the the main schedule. So it's pretty it's pretty clever. It understands numbers. So anything that you wanted to to measure or optimize for that can be numeric, we can probably nail that.
00:14:43
Speaker
The space of AI schedule optimization is, I guess, hot slash maybe a bit hypey. And I think there appears to be, if i'm if I'm not mistaken, some fundamental difference in approaches. I guess I've heard other providers talking about the importance of the network and the link graph, task descriptions and understanding what tasks mean, and then the trend of the tasks over time on versions of schedules as the task as the project progresses. But it seems as though Alice has a broader interest in resources and cost. And do you think that's, is that like a unique approach?
00:15:22
Speaker
Yeah, so it definitely is unique. We're yet to find, or someone chasing after our tales in a serious way, which is very cool. But yeah, I guess the whole thing about a network, if we take any schedule that's being planned in legacy tools,
00:15:40
Speaker
If we're building any kind of insights on top of that, we're assuming that network represents reality of you know one or a few planets that have put that together. And they're not they're not trying they're not trying to misrepresent reality, but realistically is every link they've got there the rule.
00:15:57
Speaker
And I would say highly likely that it's not. So I guess our approach at Alice, it's more flexible that this network can and it will change because we've seen so many projects and that plan they started with at the beginning is not the plan.
00:16:12
Speaker
at the end, or the plan at the end is just an artifact of a contractual agreement and they have everything else in, you know, a Excel or whatever other tool they're using to manage it. and That's why, because I mean, I don't 100% believe in those networks, why build on top of them? You know, why not actually try to represent how projects are evolving, which is when something happens on site,
00:16:36
Speaker
How do we respond to this? Like how are we going to change the sequencing, add people, et cetera, which is what you simulate in Alice as opposed to here's some contractual plan. Are we actually going to go and like re August? No, we'll probably just not tell the owner as long as we hit the milestones. They don't need to know.
00:16:51
Speaker
the YBase kind of insights on that. I guess the same with naming. it's It could be anything, right? And and I know large ah natural language processing helps a lot with naming, so we can get natural understanding of names. And it can be helpful, but overall, it's not important. It's not something you need to bring in to Alice. You can name things what whatever you want, or if you want your project to be really secret, you can use code names.
00:17:18
Speaker
And trends over time, we do actually use a bit of
Real-Time Project Management with Alice
00:17:22
Speaker
that. So with clients where they are using Alice in execution phase, for example, some of the big rail projects in the UK, what they might do is they've got their Alice model. And then they say, oh, excavation was a mess last week. It rained a lot. We didn't hit any of what we were aiming to. So what they can do is put their rate that they did hit and imagine, OK, if we continue at this production rate, where are we going to end up? If we are able to pick up?
00:17:48
Speaker
If we're faster, slower, if we add excavators, if we do over time, what would that mean if we propagated that throughout the schedule? And that's like a two minute kind of scenario and a few clicks to understand that. And so for those, for those projects, i'm I'm really interested to understand whether the majority of them are taking that first of the two approaches you described, which is like the kind of building it around rules, uh, or whether they're doing the, here's the latest version of our P six schedule. What do we do now? Sort of thing.
00:18:16
Speaker
yeah but Like, here's the latest version, help. And then... what what's What's the more common approach? The more common approach is, here's our P6 help. And only our more seasoned clients that have been with Alice for a longer time are like, let's go you know blank canvas. like Let's really yeah explore the solution space. And it depends whether they also have the power to implement what they're going to come up with. So if they're an owner, usually they can, right? Because they're the boss. So they're like, let's go for the blank canvas. Let's push this to the limit.
00:18:50
Speaker
If they're a contractor, you know, they've got interfacing lots, they've got some other constraints. They want to start with what they have and then make incremental edits and optimizations. So we really see both sides and it depends on the contract, who the- And the phase, the phase of the project and stuff like that I'd imagine. You could end up in some pretty interesting
Collaboration Between Contractors and Owners
00:19:10
Speaker
situations if like a contractor has a schedule and that That's the plan they're working to. The owner runs the scenario and says, you could be doing it way quicker. And then the contractor must say, cool, instruct me. Right. And then you get a price. Like, yeah. So actually what happened with both sides. Yeah. So some, we actually have some very cool clients that are using it together, which is rare, but I'm like, wow, this is utopia. It does exist. And then we have, you know, people looking at their own interests, which is pretty common if you bought the software for yourself. Often what ah how we can help owners, for example, is an owner might have a change. You know, like they were responsible for giving site access on a date and they can't. For whatever reason, the planning consent takes two months longer. And then they tell the contractor and the contractor goes, oh, you know, my people, my equipment's going to be sitting around for two months. This isn't.
00:19:59
Speaker
You know, here's the 2 million kind of change request or variance, whatever you want to call it. And what the owner can do in Alice is say, okay, I'm going to implement this two month delay, but what else could they have done actually just with this new constraint? So it gives them a bit of a pushback as opposed to this ah information asymmetry of like where the contractor and this is what it is. And you need to pay to actually, did you think about doing this and that based on this change?
00:20:26
Speaker
i think It's an incredibly direct return on investment. If you can get that change down from 2 million to something else, it's probably paid for others many times over. Yeah, that's really powerful. going back to the Going back to some of the constraints and this idea of almost like generating from rules, um and you talked about how task names are not that important.
Resource and Task Management in Alice
00:20:50
Speaker
how how does it if i If we're building a set of rules to generate this kind of ah scenario,
00:20:56
Speaker
How is it getting to an answer of like this should be 20 operatives or do I guess the question that I'm trying to ask and not and struggling for words to get to is is how how good question so if we think of a resource pool and gangs that can either be something that's flexible or limited right like we could fix a pool of we have 20 electrical.
00:21:18
Speaker
gangs, that's it. And so that then becomes a constraint that Alice will never have, you know, 20 of those gangs working concurrently, or the maximum will be 20. If it's free, and we just say we have electrical gangs, how is Alice coming up with that number? What Alice will do, it it makes every, jury you can make every duration parametric. And so by that, I mean, if this is a hundred hours of work, it's a hundred hours of work. If there's one crew, if there's two, it's going to take half the amount of time.
00:21:48
Speaker
Well, Alice is going to do it. It's going to try it with one, two, three, four. It's going to go up and run it with all the different amounts and the duration will flex. And it will then suggest the number of crews ah that will speed up the critical path the most as one of the solutions, as the fastest solution. What it won't do is add 20,000 electrical crews, because as soon as it finds that eight is the optimum, and if you add nine,
00:22:14
Speaker
ah the subsequent works is what's driving it, not electrical works. It will never add that ninth crew, because by default, that algorithm looks for duration and then cost. So I don't want to add more costs, so no point in adding extra beyond eight electrical crews. So then it will come to eight. How is that rule being defined? like and That's all built on top of that rule that you talked about right at the start, which is if it's one crew, it'll be this long, two crews, this long three.
00:22:43
Speaker
Is the algorithm helping to find that or is that ah is that one of the rules that you kind of have to bake in at the start? Yeah, so it's a rule you can bake in at the start. So it might not be the case that adding more crews speeds up this work, you know? Like if we were renovating a room and we start to put too many people, too many painters in there, actually that productivity doesn't get multiplied. ah The task durations you actually can state as a formula. So instead of having a number, it's like the quantity,
00:23:11
Speaker
divided by a production rate, divided by the number of crews, or maybe it's not linear. And you can just bake that into the formula as you would in any kind of mathematical formula, and it will calculate as you change those variables. The the accuracy of the outputs or scenarios seems to be so hinged on production rates.
00:23:30
Speaker
you Do you produce the production rates or are they taken from P6 or does the contractor control them? Because it seems like such a big variable that if the production rates out the scenarios like could be wildly wrong. or Yeah so this isn't reading from a big database and suggesting a production rate. It's up to the team using it to define those rates or create a library or bring in their library and it's then more of a walled garden. So they play with their own rates, tweak it as they want push it up, push it down 5%, 10% and see the impacts. But yeah, it is dependent on the rules you give it. And we're giving, the humans are defining the rules and humans can make mistakes, right? Which is also why the human then needs to review the output and go like, yeah, I believe this. Because often in reviewing it, you're like, ooh.
00:24:17
Speaker
I am building the foundation before I'm excavating. I forgot to put in that constraint. And you go back. So it's quite iterative. It's unlikely that you run one scenario in Ellis and it's you've nailed it. You've got the perfect project. It's you go check it. Oh, actually, this is where the kind of critical work is. Let's focus on parametricizing just that section and see where we get to.
00:24:40
Speaker
so it's exercising that's a summit thats That's 100% the biggest word ever said on this podcast ever. and Once you learn German, Jake, then you'll have longer words.
00:24:52
Speaker
yeah yeah um I think so. it's it's like a It's like AI for construction scheduling without the black box. I like that. Thank you.
Alice vs Traditional Planning Methods
00:25:05
Speaker
But yeah, there is definitely like, it doesn't replace humans. Actually, we could automate more of the product, but I think we think that will make it more black boxy. There is this degree of I want to have control of some of these inputs and some of these rules. And I want to review the outputs. I don't just want one output that the AI has picked for me because I haven't trust the process.
00:25:27
Speaker
So it's like kind of being given a piece of like black box is when you're given a piece of cake, maybe you're gluten-free and you're like, oh, I don't know. But this one, you've seen the ingredients that went in and you're like, yeah, it's like the old school, non-school analogy for you guys.
00:25:45
Speaker
but We touched on it earlier, and I'll probably get bad feedback from planners for assuming this, but I'm sure you could go away and spend a week on like a what-if scenario right in P6 just to see the outcome of ah an assumption or ah a change in method. I guess you're just you're getting 200 answers in a click.
00:26:00
Speaker
And then you can make your own decision based on that rule set. say yeah Yeah. And actually we don't give, we try not to give too many answers in a click. So what and the second part of the algorithm is optimizing on itself. So whenever it produces its first solution, it's then trying to beat that solution.
00:26:16
Speaker
and it will only show kind of like the top six and it will truncate the rest, so you don't have to just you know mine through everything. and All the solutions are showing on the space of cost and time, so you can quickly go like, this one's faster, this one's more expensive. and Just this week, we will release the feature to change those axes as well. so If instead you wanted to look at crew utilization, and of like productivity time on tools versus time,
00:26:42
Speaker
or cost, if those metrics are more important, you can flick between them. So you might have a bunch of scenarios, but quite visually, straight away, you can tell we're aiming for this corner. Let's pick the one in the cheap fast corner or high productivity fast corner, for example.
00:27:00
Speaker
George there's a chap I met with on a project a number of months ago and he uses, he describes that in his world there there are two types of projects. There are production based projects where they're doing a lot of things repeatedly and the game is about like quicker cycles, better production rates. And then there are like planning based projects where everything's a little bit unique. You're doing one thing and then something almost totally different next to it.
00:27:26
Speaker
Given that the underlying rule set is very production based in what we've described so far, is there a a certain type of project that's better suited to to Alice? I would actually say the more complex and bespoke it is, the better, because it's harder for a human to comprehend the variables of something that is more planning based.
00:27:49
Speaker
If it's production and cycle, for sure, the AI will smash it. But maybe you can also glean you know where it's going just because it's so repetitive. You have to produce more segments. like We'll be faster. like It's a bit more simple. You could simplify it. But with stuff that is you know industrial, with construction work packages and systems, and every package is different, I would say impossible for a human to pick the optimum solution. So yeah throw AI at it. Let it crunch away.
00:28:19
Speaker
And very, well I'd imagine there's quite a lot of work, not in like a negative way, but there's like quite a lot of work required in making sure that in those answers rules aren't missed, given the bespoke complex nature of the.
00:28:31
Speaker
Yeah, for sure. So there's i mean due diligence in traditional planning as there is with whatever comes out of AI planning. So we try and make it easy for people to review and compare options. So I spoke about you can import a P6 and start from there. So what we do, that schedule will actually show in our solution space. So then you can see a direct comparison, kind of Gantt overlay key metrics that have changed.
00:28:58
Speaker
variants between durations all pretty quickly. So that can help people believe and understand, you know, what has happened. And is this still ah kind of in line with what I wanted? Or like, where are those changes?
Future Strategies and Conclusion
00:29:12
Speaker
And like, let's just zoom in and validate that rather than Yeah, spot like the commissioning's come forward six months, how the hell that can't be possible or something we might have, we might need to double check that or something.
00:29:23
Speaker
Yeah, a bit of common sense, but definitely there's room for planners to review and verify their schedules rather than kind of spend time sweating, coming up with all the different scenarios and options over months and months.
00:29:36
Speaker
Yeah, that's really interesting. i think there's i can I'm seeing this future probably being a little bit selfish and thinking about product strategy. We we talked a lot recently internally about how ah what we should be doing more of is helping ah contractors update their master schedule. And if it sounds like the handoff to Alice's each month or whatever their cycle is,
00:29:57
Speaker
putting the the new version in and and finding out what the best optimized longer-term strategy is for them. You can see a ah better way for them in what currently is a pretty manual process and the planner probably still sits the heart of the whole thing. They're just getting rid of a lot of quite manual handoffs and crunching and running what-if schedules.
00:30:18
Speaker
Yes, you're sure. I think we're out of time, so I'm going to wrap it up there. But I probably could talk for another hour and a half, Georgia. um So everyone, thank you very much for listening. If you are interested in what we just covered, I definitely recommend checking out Alice Technologies AI for construction scheduling, as I discovered with our AI on the home page. But yes, super super interesting. ah Thank you very much for listening, and see you all in the next one. Thanks. Thanks, Georgia. Cheers.