image: thumbnail.jpg
Bringing data science to the construction industry | Blake Abbenante | Data Science Hangout
We were recently joined by Blake Abbenante, Director of Analytics and Data Science at Suffolk Construction, to chat about his career journey in data science, implementing modern data practices in the construction industry, innovative applications of AI and data science in construction, and building a data-driven culture in a traditionally less tech-focused sector.
In this Hangout, we explore innovative applications of AI and data science in construction. Blake shared how Suffolk Construction is leveraging cutting-edge technologies like AI to revolutionize traditional processes. One focus is their GenAI scheduling tool, which aims to augment and speed up the design and planning phases of building projects. This tool has the potential to significantly reduce the time planners spend on creating schedules, moving from weeks to potentially minutes or hours for an 80% completion rate. Blake discussed the development and implementation of safety models that forecast risk on projects, enabling proactive measures to ensure safer construction sites by predicting which projects might require additional safety personnel based on historical data.
Resources mentioned in the video and zoom chat:
🔗 The ellmer R package → https://ellmer.tidyverse.org/
🔗 The chatlas R package → https://github.com/posit-dev/chatlas
🔗 Posit Blog Post on ellmer → https://posit.co/blog/announcing-ellmer/
If you didn’t join live, one great discussion you missed from the zoom chat was about the challenges of data collection and analysis when encountering pushback from those whose work is being analyzed, and strategies to build trust and demonstrate value. Let us know below if you’d like to hear more about this topic!
â–º Subscribe to Our Channel Here: https://bit.ly/2TzgcOu
Follow Us Here:
Website: https://www.posit.co
Hangout: https://pos.it/dsh
LinkedIn: https://www.linkedin.com/company/posit-software
Bluesky: https://bsky.app/profile/posit.co
Thanks for hanging out with us! 💛
Transcript#
This transcript was generated automatically and may contain errors.
Welcome back to the Data Science Hangout, everybody. If we have not met yet, I'm Libby. Hello, I'm a data science community manager here at Posit. I'm also a data science educator with experience teaching both Python and R, and I am in San Antonio, Texas. If you don't happen to be familiar with Posit, Posit builds enterprise solutions and open source tools for people who do data science with R and Python. We are also the company formerly known as RStudio, so if you know RStudio, you know Posit.
I am joined by the creator of the Hangout and my co-host today, who is behind the scenes in the chat. Rachel, would you like to introduce yourself? Sure. Hi, everybody. Nice to see you today. I'm Rachel. I lead customer marketing at Posit, and as Libby said, I'll be behind the scenes here in the chat, but helping to gather the questions as well. Thanks for joining us.
Yeah, the Hangout is our open space to hear what's going on in the world of data across different industries. We get to chat about data science leadership and connect with other people who are facing similar things to us. We get together every Thursday, same time, same place, so please, please, please come and join us live if you are watching on YouTube, and if you are, there's information in the description for you to join and put it on your calendar.
So, I want to thank everybody. You have all made this the wonderful, welcoming, friendly space that it is today, and we're very committed to keeping it that way, so if you have any feedback about your experience, there's going to be a Zoom survey. It's going to pop up in your browser after you leave this meeting. I think even if you leave halfway, it'll pop up. Please fill that out. Let us know how we're doing. We would love to hear from you, but you can also always find us on LinkedIn or Blue Sky. I think that we are the only Rachel Dempsey and Libby Heron on Blue Sky and LinkedIn, so we're pretty easy to find.
At the Hangout, we love hearing from you. It doesn't matter what your years of experience are, what your title is, what your industry is, what languages you do or do not use. We really encourage you to hang out in the chat together. Introduce yourself. Say hello to everybody. What's your role? Where are you based? What do you like to do for fun? I feel like we always really connect in the chat over things that we do for fun. I think a couple of weeks ago, there was a fantastic running thread that was going on with some really amazing advice on distance running. So, yes, get together, and then also, if you have a role that you're hiring for, please share that. If there's a role on your team or that you're aware of, or if you are looking for a role, please be very vocal about that. Let's see if we can connect people in our community with roles.
All right, this is usually where I say there are three ways to ask questions, but we're going to try something different to see if we can streamline questions today. We are going to ask questions in Slido only, and let's just try this, right? So, there's going to be a Slido link in the chat. I'm going to try to make sure that we share this several times because we know that people come in and out of the chat or come in and out. So, in Slido, if you are not familiar with it, it is anonymous, but you can put your name there. So, put your name next to your question, and then there's also the ability to upvote questions. So, take a look at Slido. See if there's a question that somebody else has asked that you really like. You can maybe, like, move it to the top of our queue for consideration, and, of course, if you want to be anonymous still, absolutely, that's fine. And if today you don't have a mic going, I think it should work if you just put an asterisk next to your name in Slido or somewhere before or after your name. That way, if you have a question that you can't ask live, I can't pull you in. We can still read it for you, okay? All right, let's see how this goes today.
I am so excited to be joined by our co-host, our featured leader today, Blake Abbenante, Director of Analytics and Data Science at Suffolk Construction. Blake, would you like to introduce yourself? Tell us a little bit about you, what you do, and what you do outside of work for fun.
Sure, thank you. I'm excited to be here and talk to everybody. So, as Libby mentioned, my name is Blake Abbenante. I am Director of Data Analytics and Data Science at Suffolk Construction, DADS for short, or at least that's what I kind of try to get going, but it has not been widely adopted yet as an acronym. I've been here about two years, and we are kind of like – so, Suffolk Construction is a general contractor, which means we do, you know, primarily construction management, and we are kind of a central analytics and data science function, so we support the entire org. There's about 3,000 employees. The vast majority are kind of in the field on construction sites, so most of our, you know, support is for kind of the corporate entities, so, you know, all the kind of common ones you would have at other organizations, HR, people and culture, we call it here, finance, et cetera, and then operations, which are people in the field supporting them.
So, I lead the team responsible for analytics, which mostly in this context means kind of like cross-functional strategic analysis, and then data science, which is, you know, what you would imagine data science to be, like building products with data and to be optimized in terms of how we perform construction duties. Awesome, and for fun outside of work?
Yeah, sure. So, for fun, you know, I probably comment to like what most people say. I have two kids, a nine-year-old and four-year-old, so that is a lot of time spent with them, you know, primarily. We live in Boston, but we tend to be like more of a summer family, so sort of hibernation mode now, and in the summer, spend most of our times at the beach. I'm trying to, you know, I surfed a lot more when I was younger, trying to get back into it. Started last summer with my daughter, and my dream is to like get my daughter and my son all surfing together at some point in the future, so that is what we do in the summer. In the winter, I am kind of a big Celtics fan. I've been a season ticket holder since 2006, I think, so spent a lot of time back and forth with the Celtics games and going to the Boston Garden.
Awesome. I've never surfed. I think that I'd be terrible at it, but it sounds like a lot of fun. Well, it's like, I used to like snowboarding, like, you know, the falling on the surfboard is generally better than the falling on a snowboard, so. I bet, yeah, unless you have like really, really nice powder, I guess. I've never snowboarded either, but my husband is a snowboarder.
Types of problems and data
Well, I was wondering if you could give us a good example of sort of the type of problem that your team works on. You don't have to be super specific if you don't want to or you can't be, but I think it would be helpful to understand the types of problems that you work on and also the type of data that you have access to use. That way we can sort of form some context around this, because we come from all different industries and we won't all know exactly what you do.
So yeah, we support, like I said, you know, some of it would be things you'd see everywhere. So, you know, supporting HR, you know, we do a lot of churn and attrition analysis, forecasting for hiring. Some of the construction specific type stuff that we focus on then is, you know, safety is paramount. So predictive models to understand, you know, relative safety on projects. When do we think projects will be at a higher risk based on, you know, previous factors about just the building itself, but, you know, the work that has led up to that point in time. And then a lot of stuff around, you know, resources, procurement, like how do we plan for projects effectively? Because if you think about, you know, especially for the GC part of the business, you know, there's a lot of stuff that goes in planning. And then once a shovel goes in the ground, you're kind of set. Like, so you kind of plan accordingly. You make a pitch to an owner about what you're going to build for.
So, like, as opposed to, like, usually when you do construction at your, you know, at your house, where it's time and material. So if a contractor goes on forever, you kind of pay for it. That is not how, you know, commercial construction generally works. You kind of have a guaranteed maximum price, and that's what the owner is going to pay. So you want to make sure you plan accordingly and, you know, get a price that an owner will agree with, but also leaves you as much, you know, potential profit as possible. So a lot of support on the planning before a construction project starts, and then understanding, you know, the timeline and how we think we're trending towards, you know, not eating away at that potential profit line once a shovel goes in the ground.
Okay. I was thinking when you said, I feel like you said safety in there, and I was like safety, like physical safety, like construction safety. Are there two aspects to this, like risk safety and also physical safety? Yeah, so when we say it's safety, we generally mean, like, on-site safety. So, like, you know, if there's going to be an incident, but then there is safety, like, you know, just insurance as well. That, which we generally think of more of insurance, but I guess the insurance and risk, but there is, like, a safety, I guess, a safety aspect to that as well.
Okay. Yeah, I mean, I feel like construction is probably one of the only places where analytics or data science would really go into physical human safety. In any way, I can't think of a lot of other industries besides maybe emergency services, right? Emergency responders that would be in that zone. That's really interesting.
I feel like construction is probably one of the only places where analytics or data science would really go into physical human safety.
Construction industry and data maturity
I also am kind of wondering, I feel like I've heard you say that the construction industry is just sort of, do you feel like it's behind in analytics in general? Yeah. Why do you think that is? Even more broadly than analytics, like if you, I mean, you know, there's been multiple studies done by, like, McKinsey and other places that, you know, a lot of the optimization and efficiencies that have been gained as computers have, like, worked their way into other industries have not been realized in the construction industry. So, you know, I think the typical construction project runs 20% late and 80% over budget. And yeah, so it's due to a lot of factors.
Like, if you look at the average, you know, I think one of the common ones people also always refer to, like, Empire State Building was built in a year and 30 days or something. And, you know, building that size project today would, at best, you probably finish in maybe five years. So, you know, there's a lot of reasons for that. I mean, well, one of it is safety. Like, there's, you know, fatalities in the Empire State Building and people are much more cognizant of safety. But I think generally the industry has been slow to adopt, even like standard stuff. Like, there's a lot of smaller, you know, construction firms that using pivot tables in Excel is almost like a bridge too far or, like, is, like, the edge of, like, how they interact with data. So, that's why, you know, some of the larger ones are definitely making forays into it. And then there is, like, some, you know, construction tech. So, like, cool automation stuff. But, like, the actual management of a project and, you know, just how you think about consuming data and using data to make decisions has been very, very lacking. And I think some of it is just the logistics, like, right? Like, people on site, like, the last thing they want to do is pull up a laptop and, you know, and start looking at data that way. So, but also I think just the tool, like, there's a lot of antiquated, you know, digital tools that, and they don't kind of easily talk together either.
Right. Jared in the comments just said, my dad works in construction and still gets me to help with his Excel tables. I think if you're a data person, you have either sit there, you're the person in everyone's life that helps them with Excel tables. Yeah.
Tech stack at Suffolk
Well, speaking of tech, what does your tech stack look like? Like, what did it look like when you came on board at Suffolk and have you sort of influenced it to change it to try to, like, modernize it and bring more in? I mean, I will say, you know, again, like, relatively speaking, I was, when I came to Suffolk, you know, and talking and interviewing here, I was really, I was surprised, like, you know, I saw the job posting. I think I saw it one or two times. I was like, I don't know. It doesn't seem like, I'm not so sure. And then I applied and I interviewed. And at the time I interviewed my boss at that time was, so she's, she's chief data officer, so I was already surprised that they had, like, you know, someone at the, at the CEO, at the executive level with data in their title.
So, and then, you know, talking to her, we, they even had, like, a data engineering team, like, you know, a, you know, five to six person data engineering team, which was surprised, like, it's bigger than other data engineering team places I've been that were, you know, supposedly data and digital focused. So they had invested quite heavily in resources. Most of the data to that point, or, you know, how it manifested itself was data engineering for consuming data across all these disparate systems and mostly, like, data visualization. So we had, like, a Tableau development team.
So when I came in, in terms of, like, data science, like, most of the work that we had done had been, it had grown very organically and there was no, like, enterprise grade kind of infrastructure to build it. Like, we had, you know, one of our workflows was someone had, like, uploaded an Excel sheet to, like, SharePoint, data engineering ingested into SharePoint and pushed it down to Redshift and they ran an Alteryx flow that then, on a desktop, that wrote it back to, like, another Excel file that was then ingested into, like, another data platform. And so, you know, it was good work that I was doing. There was just, like, everything was different and kind of very non-transparent.
So when I came aboard, that's where I was, you know, so we run Databricks as a backend and we had just started to set up Databricks as a data warehouse, but the rest of our kind of outward-facing tools were lagging. So being a fan of Posit, it seemed like a good synergy for, you know, having a platform that would allow us to make our analysis mostly focused to start with, like, our analysis efforts more transparent and reproducible. So we started with Workbench, quickly moved to incorporating Connect as well so we could, you know, the work that we did in Workbench, we could publish and have kind of, like, a repeatable workflow. So we've stood that, you know, so from a tech stack, we have Databricks and the data science side, we do most of our stuff in Workbench and Connect. We also, obviously, I mentioned we do Tableau for kind of standard dashboard visualizations. We also are just, like, Microsoft is very big in the construction industry, so we're kind of, like, outside of the data team, we're kind of a Microsoft shop, so there's, like, some Power BI dashboards floating around as well. And then we've kind of delineated the line that's, like, anything that's, like, not static, we build either, like, you know, we build Shiny or Streamlit apps through Posit, and deploy through Posit Connect.
And we're starting to move towards introducing a semantic layer, so we're, like, rolling out Cube, so we kind of have a standard interface for our metrics across the organization, because that's been another one of the challenges, like, even not, you know, unique to us, where, you know, maybe the metric that you see on Power BI that's supposed to be the same on Tableau is slightly different, so you spend the first 20 minutes of a 30-minute meeting under, like, trying to figure out what the nuance in that is, so we're trying to move all of that kind of logic as far back upstream as possible, and, you know, kind of build out a more formal structure of a data model in Databricks, and then use Cube to be, like, the interface for really all consumption layers.
Deciding when to use Posit versus BI tools
Awesome. I have a, there's a question that came in anonymously on Slido that really ties into this, so I figured I would take an opportunity to ask it now, which is, how do you decide what to use Posit for, I'm guessing, like, Posit products, Workbench, Connect, stuff like that, and when to use those BI tools that you have access to?
Yeah, so, I mean, I will say, like, what I wish would happen, and then, like, what actually probably happens, which is probably, like, my hottest data take, which is, I mean, it's, like, two-sided. One is that, like, people, most orgs move, like, way too quickly to dashboards, because people say they want to get involved in data, and, like, the most tangible way that you think about data is a dashboard, so, you know, you end up building a lot of dashboards that once people start using them, the value is maybe not there, so I always prefer to, like, do more kind of bespoke analysis, and then when you get to the point of understanding what is really important to the business, like, that is what you codify and put into a dashboard, so you have, like, fewer dashboards, but they're more meaningful, because they kind of measure the things that are more important to the business, so that is how, like, I would draw the line.
I mean, I feel like you need to meet people, you need to meet your business stakeholders, like, where they are, like, just all of a sudden standing up a bunch of dashboards, especially if they have, like, no context, you know, data, I mean, construction, Suffolk in particular being a really good example that, you know, people went from, like, no dashboards to all of a sudden, you know, three dozen Tableau dashboards, and then just trying to navigate their way through that is not super helpful, so, you know, sometimes things that are in a dashboard, if it's, like, you know, if there is no need to filter and the data is not that dynamic, maybe you just do, like, exception reportings that you send through email, so, like, we're trying to move that kind of stuff to Posit, right, like, does it need to be a dashboard, do you need to interact with it, or is it just I want to get a notification when this thing changes below a certain threshold, and, like, so we look to, like, move and automate that stuff through, you know, Quarto and emails and scheduling through Connect, so that is one aspect, and then kind of, like, more broadly, our guidelines are, like, if it's just tabular static data that people want to pivot and filter, like, that would be a BI dashboard, but if it's something that, like, you need more interactive calculations or, you know, any kind of interacting with a model or other API, we usually look to build something, you know, custom through and deploy through Connect.
most orgs move, like, way too quickly to dashboards, because people say they want to get involved in data, and, like, the most tangible way that you think about data is a dashboard, so, you know, you end up building a lot of dashboards that once people start using them, the value is maybe not there, so I always prefer to, like, do more kind of bespoke analysis, and then when you get to the point of understanding what is really important to the business, like, that is what you codify and put into a dashboard.
Awesome, yeah, lots of, lots more flexibility when you can do it that way, for sure.
Historical data challenges in construction
Well, I have some amazing Slido questions that are getting upvoted. Zach asked a great question. It's got asterisks next to it, so I'm going to ask it for him, and it says the Empire State Building didn't have a lot of precedent for it. That's very true, and I hadn't thought about that. Does the construction industry have issues with a lack of historical data for new projects that are really ambitious like that, I'm guessing?
Yeah, so, I mean, even, like, you know, the Empire State Building itself, like, even given that, you know, the, there's the construction project itself, like, how complex is the building, but then as more buildings get built, like, get built, like, even just where you're working becomes, you know, the challenges of that are compounded as well. So, I think, I forget what someone, I was talking to someone here, one of our estimators, and I forget, I want, I don't want to say what the number is exactly, but, you know, building in Boston, about, like, 40 percent of the budget is, you know, just moving materials around or getting in or off site because the access is so limited. So, you know, there's, and there's no data for that, like, to account for that, but, like, even to your basic question, yeah, there's definitely a lot of challenge with historical data.
You know, we generally think we're pretty good about data relative to the rest of the construction industry, and we do have a lot of it, but, you know, some of the tools, especially if you think about scheduled data, they use, it's a tool that's now owned by Oracle. Primavera P6 is, like, the, like, go-to, like, and that is, you know, it's owned by Oracle, like, most Oracle stuff. It doesn't seem like it's been touched in 25 or 30 years. So, there's not a lot of great ways to get data out. It's kind of very messy, and then how people set up their projects are always different. So, like, looking for, like, a through line between a project that was stood up 25 years ago and how, you know, a project is built today, there's not a lot of continuity in that data.
Yeah. I guess that's the same problem that everybody has, is that historical trends don't always equal future projections, right? And once we got past COVID, I think we all realized that that was true in a really big way, that nothing would ever be the same again. And I feel like COVID, well, and also just supply chain things that were happening during that time, really changed construction with construction material costs. Is that something that you had to deal with or the aftermath of that when you came on? Because you came on to Suffolk in 2023, right?
Yeah. So, we're, like, so it's, you know, right on the heels of COVID, basically. But all the projects that took place then, I mean, some of them, I mean, we actually have done a good job. Like, they actually went back and, you know, coded a lot of the stuff. Because there's even just some costs that were, like, additional costs that were associated with projects or additional PPE that they needed during the COVID time that, you know, all else being equal would even then skew the data about how that project performed relative to the before time. So, we've done a good job of, like, trying to identify some of that and excluding it. But even just more broadly, like, yeah, like, the second or third order effects are, you know, not as explicit and much harder to identify.
I will say, like, the procurement and supply chain, it obviously affects us. Suffolk as a GC, generally speaking, is somewhat shielded from those costs. Like, they're either, you know, on the owner or on the subcontractor. But it does affect, like, the overall schedule. So, you know, it's not something that we can completely ignore. But, yeah, like, that is always a constant. All right, people want to look at historical data. Like, you just put a big asterisk for the three years that we were doing, you know, COVID-related projects and carve them out or, like, just make note of, like, how they change it. And we don't have, like, a hard and fast rule for how we do it. It really – I hate to be, like, you know, you kind of look at it and make a judgment. But, you know, that ultimately is – because, like, that's the other thing, too. Like, you know, construction, we're rather large. But if you think about the unit of measure being a project itself, you know, and at the scale that we usually do most of our analysis, there's, you know, maybe only 30 or 40, maybe a part of 60, you know, projects that finish in a given year, right? The lifecycle of a project being, like, three years that, you know, at that level, it's kind of small sample sizes all over the place.
Oh, wow. That's such a long time horizon because inflation changes a ton even over three years. All right. Well, enough of my nosy inflation questions. There's a lot of really good questions on Slido. Reminder for everybody, if you've just hopped on, we're asking questions in Slido today, which is actually pretty cool because you get to upvote stuff. And I'm loving seeing what kind of rises to the top. So, we have an anonymous question that has risen to the top of our Slido.
Getting buy-in from site managers
And it says, do you get pushback from site managers who place more trust in their subjective experience than what the data is telling you? And if so, how do you get buy-in from those people?
Yes, 100%. And, you know, I wouldn't even say, like, I don't know if I frame it as pushback. A lot of it is just, like, they know better, right? Like, you know, there's nuance in the actual management of a project that is not captured in the data that we have available. So, generally, what we want to do before we roll anything out is kind of, you know, identify site managers, superintendents, PXs, people who have more intimate knowledge of projects themselves to get their insight and feedback on what we're doing before we go roll something out.
And I will say, I hope we've done a better job since I've started, but that has, I think, been, again, probably not specific to Suffolk. But one of our big challenges here is that, you know, we've done a lot of good technical work, like the 80% of getting the technical work done, and then we've really dropped the ball on the last 20% work, which is, like, you know, what's the productization of this? Like, how do you get end users in a way that it makes sense to them? Some of it is as simple as, like, how you label it. Like, when we rolled out, like, you know, the example I gave before where we called it, like, the safety model, it somehow morphed into, like, and the way the data was presented, like, it was kind of like you have a higher likelihood of incident. So, it turned in from, it turned, the initial rollout turned from something being, you know, helpful or useful to, like, a list you wanted to avoid, and that actually would impact how people collected data and, you know, tuned out the results just by the way that we framed up with what the solution was.
So, you know, a lot of lessons that we've learned there, which is the technology itself, whether, you know, the model building itself is, you know, maybe at most, like, 50% of the problem. The rest of it is, like, how do you roll it out and make it something that, you know, communicate with your end users so they understand what you're doing and how it can best, you know, how it can best make them more efficient at their job. It's not something that we're trying to use data to be big brother and flag people for things that are beyond their control. So, yeah. So, I mean, that is, like, a constant evolution. So, like, that is, broadly speaking, is, like, we do try to do a better job of partnering with our strategic business stakeholders so they can be part of, like, how we roll solutions out. But the other one is just, I mean, we're kind of going eyes wide open knowing that the data that we currently have, there's probably additional data that we would need that would give us more nuanced insights, and then partnering with, like, just people who are on the job sites to figure out how we can actually collect that level of detail that we need. So, when we iterate on these, you know, whatever the data product is, again, that we think we're improving because we've now captured additional fidelity in the data.
maybe at most, like, 50% of the problem. The rest of it is, like, how do you roll it out and make it something that, you know, communicate with your end users so they understand what you're doing and how it can best, you know, how it can best make them more efficient at their job.
AI use cases in construction
Yeah, that's awesome. Well, I have another fantastic question from Slido that has been upvoted a ton. It's from Tony. Tony, are you available to answer that? I'm going to jump in and ask. Let's see. Maybe I caught Tony off guard. I see Tony tried to unmute, but there you are. Oh, there we are. Okay. Hi, Tony. My video wasn't coming on. Yeah, I'd just like to hear a little bit about where you might be using AI in the process of budgeting or trying to figure out maybe where cost overruns are likely to happen or where you can build in to your budget to try to ensure you'll have some profit margin in the long run, or even, especially here in the Boston area, the types of customers that are more profitable, higher ed versus healthcare versus government, things like that, and which AI tools or products you're using.
Yeah, sure. Great question. So, you know, I would say, so Suffolk, one of the maybe benefits of Suffolk is we are still private. So John Fish, you know, founded Suffolk and still leads it, and he is a big proponent of data, and this past year has made kind of a corporate-wide edict. I mean, I look at the sign across the street, it says we're aiming to be the future leader of AI in the built world. So it's kind of like a company-wide initiative to figure out how we can best leverage AI. So the way that, you know, over the past year it's kind of taken shape, there are sort of two flavors of it. One is kind of what you would imagine what everyone's doing, which is, all right, we have a bunch of SOPs or PDFs somewhere, so like let's throw like those in one that we can build a bunch of chat agents that are, you know, does rag to retrieve specific information about particular domains. That one's less exciting.
But the other areas that we're looking to do it, I mean, there's kind of two large ones, and it's actually in not necessarily either of the ones that you identified, but we're doing in the design, like how can we use AI to augment and speed along design of buildings themselves, and then in the planning of projects. So can we, as I mentioned before, like the kind of planning tool that is the industry-wide standard, P6 Primavera, is very antiquated, and it takes, you know, so just from a user experience perspective, it's not very intuitive or friendly, and where we think we can like gain a lot of efficiencies is automated using AI to kind of, if not create an entire schedule, you know, take something that it may take a plan or two weeks now and get them 80% of the way there in a manner of minutes or hours even. So those are kind of some of the big areas of focus, and then kind of maybe in between area where we've seen a lot of traction is in data transformation. So like we have a lot of data in PDF, so as part of our data pipeline, can we use AI to like extract information out of that and turn it into tabular formats as part of our data warehouse? So we've picked up a lot of stuff like that. So I honestly, like in the past two or three months, like I feel like, you know, a year ago everyone was really trying to force AI and like put it in, you know, as a square peg in a bunch of round holes, and I think over the past month, like we've had like a bunch of really smart and novel use cases come up. Like literally, I think just this past week one of my colleagues built, he has an AI bot that, or an AI agent that looks at GitHub submissions. So basically like looks for any changes to our
data model and then goes and looks through all our GitHub repos and like figures out any references to it and like find out the products that are associated with it. So we can send notifications to say, you know, this change to this data model is going to have a potential impact to these products downstream. So yeah, so I think, you know, I think probably still in the short term, aside from like those two outward facing ones, a lot of the efficiencies that we're gaining are more upstream in the data pipeline side at the moment.
And then, oh, in terms of what we use. So as I mentioned, like from an enterprise agreement, we kind of have a partnership with Microsoft. So we started using, you know, the Microsoft Azure suite of open AI integration. Since then we've been, you know, we've been tasked, we've been given the ability to kind of pilot other ones as well. So Databricks has their own native LLM. So we use that somewhat. We're kind of looking or maybe potentially piloting. So we have a bunch of people using individual open AI, like business accounts, but so looking to consolidate that potentially as enterprise as well and try that. And then maybe we're, maybe again, some individuals have used Gemini as well and make, we might look to make that more and more formal as well. So really like trying to run the gamut, whatever's out there, like we're, it's kind of like the wild west still in terms of technology, like trying to figure out which one works best for which scenario.
Getting executive buy-in for data science
I think we can all relate to that. Thank you for the amazing question, Tony. Great, thanks. I have a question, and I think that Jeffrey Silverman's question on Slido kind of rolls in with mine. So Jeffrey had asked like, you know, the construction industry is hesitant to implement new tech and data science analytics. So do you find that folks at the company are resistant to your work? And I know that we had talked a little bit about like getting people to trust you. My question is, how important is it to have an executive like you've got that's championing data science efforts? And is there any advice that you have for people who are sort of in that battle right now where they're trying to get executive champions on board?
Yeah, so I will say, I mean, that is probably the biggest lesson I've learned over my career. I was like, I mean, my somewhat snarky, not snarky answer is like, you know, the best way to do it is to, when you're looking for a job, find a place that already has that infrastructure built in. Because I will say that has been like my biggest challenge in my career. Like, you know, you talk to somebody who's really about all about data, like they're the hiring manager, and then you get hired in, and maybe they're like a senior manager or director, depending on what, like, if that's as high as it goes, and then, you know, they could have buy-in, but that doesn't mean that they have buy-in from whomever is above them. So I, whenever I would interview, I would always ask like, what's the highest role of somebody who has data in their title? And like, that would kind of, to me, be a proxy for like, how bought-in are you on actually using data?
But otherwise, you know, you kind of go down the same path that you normally always do, which is like, all right, you find your strategic business partner, you like get your small wins, and then you fight for a budget. But I wish I had a better answer. But like, to me, like, there is no shortcut to it if you don't have people who, I mean, you know, you can, I've been places where you can take individual data wins to, like, people at the C-level or even right below the C-level, but unless it's like a company-wide, like, edict, like, I find it really challenging for everyone to just, for it always to be, like, top of mind and not something that will get cut as soon as, you know, those are the type of decisions that need to be made. So I wish I had a more optimistic answer for it, but like, there, like, I think, you know, they're, obviously, like, you kind of have to understand the scope of the battle that you're trying to fight. And if it is that, like, I'm a manager who's come in and, like, I only have a director above me, then, who believes in my vision, but, like, you know, then thinking you're going to be able to get a budget that to implement something, you know, enterprise-wide out of the gate is kind of, that's more like a three- or four- or five-year play, even. So you kind of have to understand, like, how high up you have approval on, and then, like, kind of fight your battles and work your plan from there, because I don't think there is a shortcut to go from, we only have buy-in at the, like, manager level all the way up to, like, C-suite buy-in. Yeah, that's not something that's going to turn around in a couple of months or even maybe a couple of years.
But I wish I had a better answer. But like, to me, like, there is no shortcut to it if you don't have people who, I mean, you know, you can, I've been places where you can take individual data wins to, like, people at the C-level or even right below the C-level, but unless it's like a company-wide, like, edict, like, I find it really challenging for everyone to just, for it always to be, like, top of mind and not something that will get cut as soon as, you know, those are the type of decisions that need to be made.
Data sharing in construction
Yeah. All right. Thank you so much for those insights and fantastic question from Jeffrey. I have an anonymous question that is next in our sort of vote on Slido reminder. We're doing questions in Slido, and you could vote on them, which is super fun watching them sort of, like, pop around my screen. So the question is, are there established practices for sharing data across businesses in construction, or do you only have to rely on your own company's data only?
Yeah. So there are no established practices. I mean, I would say even, not even across businesses, even within businesses, there's very little established practice for sharing data. So unfortunately, there's not. I mean, there are tools that, you know, construction management tools that allow, you know, multiple different people to kind of access from a operational perspective the same data, like Procore is a big one. And, you know, you can have, like, a GC like us and subcontractors, like, having, or even owners having access to that same instance. But in terms of just, like, raw data, there's not. And, like, that, like I said, even, like, there's a lot of places, especially when you get down to, like, the subcontractor level where it's all data in Excel, and then, you know, there's not even going to be a standard format as to what that Excel is.
So I don't think any of them are, but, like, I, we have, like, a great data engineering team here that, you know, thankfully, because I've been in places where you've had to do it, like, that's been a lot of the work that has been thrust upon data scientists as well to, like, all right, like, how do I, you know, standardize all these different data sources into it? So you spend, you know, 50 to 60 percent of your time just doing that. Our data engineering team is, it's great, and they handle a lot of that for us. So whether it's, you know, data that they ingest from Excel through SharePoint, then matching it up to, you know, enterprise data that we have, like, thankfully, like, you know, we partner very closely with them, but they do all of that work. So we are one of the few, I think, fortunate, lucky ones, like data, our data science team, like, we just start at the data science part and don't have to do very little data engineering.
But, yeah, you do point, like, raise a good point, which is that, like, you know, as people think about implementing AI and LLM, like, as I mentioned, like, if you're a large scale construction company and you're only doing, you know, say 30 to 40 projects a year, there is, like, a real understanding that there was, like, a lack of enough data, but there is not, like, an aggregate data broker that, like, has, you know, there's maybe a handful that have enough data construction-wide to, like, make this kind of inferences. So there's, it is a real challenge to think about, you know, do you have enough data even from your own business to, like, to do anything, like, at scale? Yeah. I mean, these are questions that I think we all sort of bump up against depending on where we are and what industry we're in.
Getting started with AI and LLMs
Well, I have so many questions on Slido that are so good. Let's see how many we can get through in the 20 minutes that we've got left. Neil Bray had a great question. It's got asterisks next to it, so I'm going to ask it. Your bio says that you have integrated AI into business processes. So for coders who are currently working outside of AI, what would you suggest as their first step into AI?
So I'll say, and this is, like, a non-paid endorsement. You know, my first foray into, like, using AI was using LangChain, and, like, I, that was just kind of a disaster for me. Like, I could, like, it doesn't, didn't really click. And then recently, like, we've done some, you know, kind of other integrations, and then over the past couple of months, using the ellmer package through Posit and also, like, the chatlas one as well, we've started both using. I think, like, makes the, like, coding to AI, like, much more approachable. So I would, like, if you have not used, like, the Posit suite of AI integrations, I think, like, that makes it all very approachable. That would be how I would recommend getting started, and it's really, it is really easy, then, to integrate it to, like, any other kind of work that you've done or if you do happen to be, you know, Posit or any of the Posit suite of tools. So building it into a Shiny app or whatever else. I think that is probably the easiest way to get started or get your toes wet.
Yeah, I see Rachel putting in the chat links to chatlas and ellmer as well, if anybody wanted to go check out. There are so many amazing packages that I feel like have just come out in the past six months or a year that have been worked on that really make interfacing with LLMs so much easier and more intuitive. MALL is also a great one, M-A-L-L, the MALL package. Really, really like that one. So go check that one out. Maybe somebody can put a link in the chat to that one. There we go. Thanks, Link. Link with the link. I love that.
Supporting Posit tools alongside enterprise infrastructure
All right. So next question from Slido. Let's see how many we can get in. Alan, Alan, you had a question. I'm guessing this is our Alan, our Alan. If you want to unmute and ask. It is your Alan.
Hey, Blake, this is a really great conversation. Thank you. I'm really curious from a kind of operational perspective how support works for the set of Posit tools you use alongside that big enterprise-y stack of Azure Databricks, Power BI. How do you carve out support there? Is that dedicated support or is it kind of shared? Does it come from members of your team? With specialized tools like that, I think it's often hard to figure out who owns those things, who makes them work in the environment, all that kind of stuff. Would love to hear how that works in your organization.
Yeah, that's a great, great question. You like kind of hit on a, I don't call it a store spot, but like we kind of almost do run two parallel environments. We have like the Microsoft environment, which is fully supported by IT. And then we have a separate AWS environment, which is where Databricks and all our Posit tools run. It's split that way prior to me being part of Suffolk. You know, we've been kind of had some inroads over the past six months or so about like, can we bring them back a little together? It doesn't necessarily make sense for us to be so separate, especially now in the world of AI, where we want to integrate some of the AI agents that we're building on this AWS side as a co-pilot. So for the Azure and all the Microsoft side, we do have support from IT. On the AWS side, kind of be careful what you wish for. We're a little on our own, but we have to support our own. We do have some outside contracted support, like MLOps or excuse me, DevOps help. But generally like, you know, I kind of administer all of our Posit tool suite at the moment, but it is my hope that over the next year that we can have more support from our IT team. I think maybe part of the biggest challenge now is like, they are, you know, like a lot of, like they are much more fluent in the Microsoft side of the world. So as we start to bring on more AWS people, like I think we have a better chance of getting more internal support. Awesome. Fantastic question, Alan. Thank you.
Data adoption in construction and HVAC
I wanted to say that there are a couple of questions in Slido that could be answered by the chat by our community. So please hop in and answer this for these anonymous askers. One is, can someone explain the difference between data science and data engineering? Great question. Another one is, what is a tech stack or a data stack? Because I asked that in the beginning, I asked Blake, like what his tech stack was. So chat, get to answering that for our beautiful anonymous askers. And our next question is from Devin. Devin, are you available to unmute and ask this one? Put you on the spot here.
There we go. Sorry about that. No, it's all right. Yeah. So I'm not quite in construction, but I work in HVAC with building automation. We actually partner with a company that offers an analytics platform to do fault detection and diagnostics. Our experience is that clients are really excited about it. I think especially getting transparency and insight where you might not have had that before, but nobody else really wants to adopt. It's too much of a challenge. And getting the infrastructure and the cost is always kind of a preventative thing. So I'm curious what your experience has been with that.
Adoption from clients in particular or just like from external folks?
I guess really throughout the industry. I mean, we see different levels of adoption at different roles.
Yeah. I mean, I think, yes. So some of the larger owners, I think, from an owner perspective, are more sophisticated and expect even pretty transparent and detailed access to data. So again, probably not surprisingly, those are the ones who have probably more automated tools that allow for the consumption to be easier. Further down, yeah. I mean, it is a little bit more of a challenge. I think where you do kind of get buy-in is that if you can draw the string between better data collection, gives us better insights, which can ultimately lead to lower insurance rates or other cost savings along the lines. Because ultimately, that's what people are like. That's going to be the bottom line is how is this going to save me money? So kind of being able to draw that line between like, all right, the better you can tell us how long people are on site, that is a direct impact on ultimately downline safety metrics. I think the more that you can get buy-in from whoever it is. I will still say that's kind of like looking up. The kind of bigger challenge is maybe down like subcontractors. Again, we work on small margins and they work
on even smaller margins. So being like you also have to collect data for us is going to be like very bottom of the line. Some of it just becomes contractual as we bring on a sub.
They have to make concessions as to what they'll provide for you. But I think the more that we can integrate the technology into the tool set that people are actually using on site versus being like a one plus thing that they have to do, or even like for late data collection is what I'm always pushing, where it's not people don't have to like explicitly do things. But if we can just like track stuff around on construction sites, whether it's through like AI or just like monitors, I think we can lower the load that we need from a adoption standpoint on the data generation side at least. Awesome. Thanks so much for asking that question, Devin.
Impact of data science at Suffolk
I have an anonymous question that has risen to the top on Slido. Reminder that you can vote on Slido, so please go check out those questions and vote for them. And it is, if you were sharing with other construction companies who aren't really using data as much today, how has the data science made an impact? So what impact has data science made at Suffolk? Yeah, so, you know, I think if you think about, like, if you think of the inverse, which is, you know, just generally in the construction industry, like most, a lot of the people are, it's an older demographic in general, right? Like a lot, like, so there's a lack of resources and it becomes, but a lot of, a lot of kind of like what you would use data for right now is done just by kind of like tribal knowledge. Like, you know, I have a superintendent who's been on site for 35 years and he knows, like, every kind of project. Those people, like, in the next five or ten years are all going to age out of the construction workforce and then you're going to be left with, like, this dearth of not having that, like, institutional knowledge.
So to me, like, that is how you, that's biggest value of data, which is, like, all right, as you kind of, you know, as these people leave your workforce and, like, they can't, you can't go to them to be, like, what happened on this project 10 years ago and, like, why, you know, what was the cost and why is it two times more than all the others? I think that's where you see the real value is, like, it creates that transparency across the org so you don't, you are not bottlenecked by, like, individuals who just kind of have the entirety of your firm's knowledge in their head. So it kind of makes the data more accessible and shareable and not siloed like that. Right, yeah, because a single point of knowledge is also a single point of failure, unfortunately, as we have all experienced, I'm sure, in our lives.
I think that's where you see the real value is, like, it creates that transparency across the org so you don't, you are not bottlenecked by, like, individuals who just kind of have the entirety of your firm's knowledge in their head.
Overcoming pushback on data collection
All right, we have another anonymous question that was, that has risen to the top. These anonymous questions, by the way, are amazing. So if you feel comfortable adding your name next time and hopping in and asking it live, I really encourage you because you are all asking amazing stuff. This one is, I was hired to collect and analyze data. When I go around to collect and analyze, I get pushback. How do I help myself in this area? I've experienced this too, whoever you are.
Yeah, I mean, like I said, I think, you know, one of the things that we, that I try to push for as much as possible is, like, can you get kind of, like, what I kind of, like, refer to as latent data collection. So, you know, is there just metadata around other tools that are already generating data that you can use to, as a proxy for it, what you're looking for? Or is there, are there more passive ways to collect data versus, like, having it be one more thing that people need to fill out and think about and do? So, you know, I think that's one respect. The second one is, you know, maybe flip the dynamic of that relationship instead of the, maybe the first time you go to ask somebody something, it's you asking them for something, you know, figure out, is there anything you could do for them first? And then kind of as a reciprocal thank you, like, then you can maybe go with your ask for data. But, yeah, I mean, to me, like, those have been the biggest pushbacks. It was, like, when I, with no context, and look, go to somebody to ask them for something, and it's something, like, totally out of their normal workflow that they didn't have to, like, set aside time and, you know, change frame of references for.
Industry-wide standardization and data
Awesome. Let's grab another question from Slido. Our next one in the voting structure is from Jared. And Jared has an asterisk, so I'll ask this for him. Following on from the data sharing question, what variables would be top of your wish list for a standardized cross-industry data set? Like, what aggregate data set would have the biggest impact on an industry-wide scale? That's a huge question, so go whichever way you would like with that, or I can re-read it for you. Yeah, no, I was looking at it. So, that's a great question, but it's a very tough one. It's a big one. Like I said, I think there may be a very easy way to think about it. Like, there's really not a standard way by which people, you know, schedule, even within Suffolk, right? Like, there's not a standard way that they define tasks and, you know, build a schedule. So, you know, when you want to compare how long did these tasks take on this schedule or this type of project versus other ones, like, having just, I mean, and there are, I will say, like, there are standards out there. It's just that people do not adhere to them. So, like, I think if there was, like, more adherence to some of the, like, kind of just industry-wide, like, how you define tasks, what needs to be done, that would propagate kind of all the way through, and then you'd get kind of more standard production rates, like how long does it take to do these tasks.
But that is the other thing as well, is that, you know, there's, like, two different ways at which time is measured. There's, you know, what people, kind of what they call construction logs, which is, like, how long are you on the construction site? And then there's, like, actual time card data, and those two things very rarely match up. So, like, understanding, like, how long it actually takes to do something is very hard to gather with, like, the tool by which most of the construction industry is used to capture that information. So, I would say it's probably those two things. Like, one is better standardization across, like, what the tasks are that we're actually doing, and then better tracking of the time it takes to do those things.
About Suffolk and other audience questions
All right, awesome. Hopefully, Jared, that answered your question, your pick-pick question. I saw Deanne Alvey asked a question probably a long time ago. It's small, quick to answer, but I do think it would be helpful. It's, for context, how large is Suffolk? How large is your firm? Oh, sure. Sorry, I thought, yeah, so we're about 3,000 people in total, and again, I would say maybe 2,500 to 2,600 of those are, like, people, like, on, you know, on site, out in construction, like, on construction site, and then, you know, the other, the remainder are kind of in corporate functions.
Awesome. All right, cool. That's great context. Thanks for asking, Deanne. I have another question here that I saw that was anonymous that I really liked. Where is it? Ah, it says, as a construction company, what data sources do you use regarding real estate values and those trends? Is that part of your, sort of, consideration?
So, it's not so much, so, it's fine. We actually, like, you know, there's Suffolk, there's multiple branches of Suffolk, and we actually, there is one that's, like, real estate investment, and, you know, in there, they probably are much more focused on that, but I will say, you know, prior to me coming to Suffolk, I did work at CBRE, and, you know, they are probably the biggest source of real estate valuations. I'm pretty sure that we use them here as well, but, yeah, like, that is not an area that, at least, my team has been involved in to date. And, to be honest, like, most of it is not, you know, we're not, like, you know, like, we're not the one who are going out procuring locations, so, usually, it's, like, an owner has already done it, and, like, they're just contracting with us to then build.
Yeah, it's such a great question, though. I mean, it's something that I hadn't even thought about, but I've also sort of been on the inside business side of that, where, like, before we've hired a construction company, we've gone through this whole process of figuring out what site or location we wanted to build or lease or whatever it is. It's a really, really interesting data problem. All right, I have a question from Rachel that has popped to the top of my slide out. Rachel, would you like to ask? Oh, it's one of my favorite questions that I like to ask people at the Hangout every week, but is there a piece of memorable career advice maybe that you've received over the course of your career so far or that you've given others that you'd like to share with us?
Career advice
You know, what I think is probably something that's common when you're kind of early on in your career, you see, like, whoever, managers and directors, like, everyone going to meetings, and you're always, like, why am I not invited to that meeting? I have, like, those are some things, and then you, at some point, like, it flips, and, like, you start getting invited to meetings, and then you end up showing up at all of them. And, like, what I always tell my team is, like, just because you're invited, like, don't, like, if you don't not have, like, an opinion, like, you know, if you're going to go to a meeting, like, show up with an opinion and, like, a purpose. Otherwise, you know, you can just get the notes and summary. So I think, you know, your time is as valuable as anybody else's time, so just don't think that because someone has sent you something, it's an obligation you have to meet. Like, you know, you can prioritize for yourself and make those decisions and be strategic about, you know, where you, where you allocate your time. Thank you. I love it.
So I think, you know, your time is as valuable as anybody else's time, so just don't think that because someone has sent you something, it's an obligation you have to meet.
Upcoming events and closing
I, I, there's a few minutes left, so I just want to sneak this in here real quickly because Blake had shared a few different examples of the ways that their team is making an impact with data science, and Max Patterson from Suffolk is actually going to be leading our workflow demo. Next month, so the last Wednesday in March, and so I just wanted to remind everybody that, and I'll share the link in the chat here, but Max will give probably, like, a 25 to 30 minute demo into one of those use cases. So he's going to talk more about that safety model that Blake mentioned and how they're using Databricks, Posit Team, Quarto, Shiny, Vetiver, everything together to bring that to life. So I'll go put that in the chat right now too. Yeah, that's awesome. Thank you. Yeah, Max is, he'll, it'll be, be really good. It's been, that's the one, like, it's a very nice evolution of that product from something that was, as I mentioned, like kind of like Frankenstein workflow to start with to like a much more sleek and business, business relevant implementation through like the Posit stack that we've stood up.
That's a pretty rare combination that we get to talk to Blake on one day. And then a few, couple of weeks later, we get like a peek behind the curtain in the same company. That's amazing. All right. I wanted to thank everybody for hanging out with us. This was amazing. Blake, thank you so much for sharing all of your thoughts and insights with us. I wanted to also let everybody know that next week we have Jessica Klein, who is a data scientist at the Census Bureau. I've just popped a link in the chat if you want to go check out Jessica's profile. I think it's going to be a lot of fun. Jessica is amazing and fun to talk to. Thank you so much, everybody. We'll see you next Thursday. Have a wonderful day.