Latest / The Punch List / Eric Lamb on why construction is its own biggest enemy

Transcript

Eric Lamb on why construction is its own biggest enemy · The Punch List

• Jon Wright (00:00:00): What's your filter these days for construction tech and making bets? Eric Lamb (00:00:04): They've gotta have a sense of what's happening in the field and not just the technology. Jon Wright (00:00:09): What does the "D" and the "P" and the "R" stand for in "DPR"? Eric Lamb (00:00:12): It stands for Doug, Peter and Ron, the first names of Doug Woods, Peter Nosler and Ron Davidowski. I've never been in our industry in a time where I've seen so many startups and so many different construction technology companies coming into our industry. Kanav Hasija (00:00:27): AI is not a magic wand. You have to fix it with the right data and the right context, Jon Wright (00:00:32): You're getting a lot of construction tech pitches, what's one that you've passed on that you still think about? Eric Lamb (00:00:37): It was Tracy of PlanGrid, they only wanted like a hundred thousand. They ended up selling to Autodesk for... Jon Wright (00:00:48): Welcome to The Punch List Podcast, the number one podcast for architecture, engineering, construction in the entire world. We have a very special guest today. TK, I'm gonna turn it over to you to introduce our guest Tanmaya Kala (00:01:06): Yeah. So today's guest is, um, Eric Lamb. He's been in the industry for 45 years, which is longer than anyone else on this podcast has been on the planet. And, um, know, out of which 35 years he served as a founding management committee member for DPR Construction. He helped build DPR from a startup to a top 10 general contractor, which did $14 billion in revenue last year. And currently, he sits in the board of DPR. He’s also part of WND Ventures. He's a partner at Blackhorn Ventures, where he's investing in the future of construction tech. So I'm really looking forward to this conversation. There is lot of history there, but also Eric's the kind of guy who's always looking to the future, so the conversation will be very great. One thing on a personal note I'll say, 13 years I was at DPR. I was very lucky to have worked with Eric. I, uh, consider him as my mentor and, um, I think one of the greatest learnings I've had in construction and personal life is through our conversations. So, um, really excited for it Eric Lamb (00:02:15): you Jon Wright (00:02:16): Welcome, Eric. The question on everyone's mind: What does the D and the P and the R stand for in DPR? Eric Lamb (00:02:24): So one of the first things the company did that was different than other construction companies. So it stands for Doug, Peter and Ron, the first names of Doug Woods, Peter Nosler and Ron Davidowski. As you know, most, uh, initials on companies have to do with last names, but they wanted to promote the familiarity and the, uh, sort of congenial and collaborative way that our company was founded. And so they used their first names and initials. I worked with them, for over 10 years prior to the founding of DPR at another small family construction company in the Bay Area. They represent to me visionaries of gentlemen that really wanted something different out of a construction company. And they, they even had their CFO as one of the three, Ron Davidowski, which was different. In those days, you wouldn't have a financial person in the group of founders, but they had a financial strength. Plus they had, uh, Doug and Peter from the construction side. But I joined them a few months after they started along with, uh, several other people Tanmaya Kala (00:03:39): Thanks for sharing that, Eric. One of the things I would say that, uh, when I had joined, we still had these guys operational, and listening to Peter talk was also very interesting because wasn't he like a physicist? Eric Lamb (00:03:50): He went to UC Berkeley and he was in, uh, the PhD program in physics and, and he worked summers as a carpenter and millworker up in Oregon where his family was. So he was a builder at heart, but he enjoyed the science and he enjoyed, the research that he did in physics, but ultimately the building part of him took over and he's joined the industry. Tanmaya Kala (00:04:15): You've seen DPR go from this small startup with few people to in our top 10 GC. What would you say was one decision that mattered more than what people might think, or what people might realize today? Eric Lamb (00:04:34): This is, um, fairly easy for me to, to, to, to consider in that as I said earlier, we wanted to be different as a construction company. And one of the things that we, had talked about very early on is what is really the purpose and what is really the importance of DPR coming into the construction industry? What's gonna be different about us? We saw Professor Jim Collins over at Stanford, in the first few months of our founding in 1990, and he talked about companies with bad mission statements and companies that lacked real, core, uh, purpose and values for their founding. And he really was very eloquent in talking about what a company needs to do when they get started, particularly in understanding sort of what do they stand for? What are their core values? And then how to develop a time-scaled mission statement that was really more measurable and do it over a definitive period of time. So we actually hired him in early in 1991 with all the employees, all 28 of them at DPR, to help us develop and, basically codify what some people would refer to as our core ideology around our core values, around, our purpose, and we did our first mission statement of 2000. That developed really, through the years as the centerpiece, the cornerstone, so to speak, of our culture and what we stood for, how we behaved as leaders, how we behaved in the industry and the community and that sort of thing, and was really, to this day, is the most important thing we ever did by far Tanmaya Kala (00:06:20): Yeah, Jim Collins' "Good to Great" is one of our favorite books. In fact, if Kanav was in his home in Berkeley, that book would be displayed right behind him over his right shoulder. Eric Lamb (00:06:30): He was still in the, first draft of that book while we were working with him, and he provided a draft to Peter Nosler to help, as one of his many colleagues that were helping him edit the book. So Peter Nosler is sort of there in spirit in some ways. Kanav Hasija (00:06:47): So Eric, I, I want you to peel, peel the layers a bit more on this question, right? This is my second venture as well, and i-in my last venture, I found the importance and value of the core values and the mission statement you guys did. But it's always understated and under-delivered, the importance of that because some people in their, uh, like the mid-sized GCs might think that's all bullshit. No one cares about core values and, m-mission statements. The real thing is money and growth and profits and revenues, right? So through some examples or anecdotes, tell us why a particular core value once iterated upon multiple times and ingrained in, every employee's mind boosts the company more than just the revenue growth? Eric Lamb (00:07:32): It is true what you described, Kanav, that a lot of companies will put a mission statement or some purpose statement on the wall, and then no one ever reads it, and no one talks about it, and they don't really live it in the company. At DPR, we talk about it all the time, and we have company meetings, and we give examples. I mean, one of our core values was ever forward, and ever forward to us is representing we're always innovating, we're always out front, we're always trying new things, we're developing ourselves and how we could be better, and from the beginning as a company, developed digital tools. In fact, this new invention called email in 1990, we adopted across the whole company. None of us had email addresses in our previous company, and none of our subcontractors had email addresses in 1990. So there was no one, nobody in the trade partners to actually, uh, email. But we did have emails with designers and owners, and that of course, developed across the industry. But we were out front in trying to encourage that because we wanted to get rid of the room with 12 fax machines, you know, going between all of our partners, in the business. I think the, idea around enjoyment, one of the four core values lived all the time. TK knows it well from, uh, many times standing around the wine bar at one of our offices. So I think if you take our four values, each of them individually, we always talk about them and show examples of it, and then we share them between everybody Kanav Hasija (00:09:16): So well said because coming from outside of the industry, and I've talked to a few contractors, when I ask them, what do you think of as emotions when you think of DPR?" And they think of innovation and trust. You know, those two are the emotions that they kind of, hear, developers or contractors when they hear of DPR. And that's why it's so important for this core values because these emulate outwards and people remember you from, for those things. So that's well said. Tanmaya Kala (00:09:47): Changing tracks a little bit here. Uh, we got to talk about data centers and advanced tech market. It's a twofold question. Obviously, with recent growth in data centers and DPR being one of the largest data center builders, Eric, one is I just wanted to get your thoughts as to how you see that market different from a traditional commercial market and, what do you see in the future in that market? I mean, is it gonna keep growing? Do you think it's gonna cool down? What are your thoughts on the future of that market? And I guess I'm gonna add one other bonus question. I think you and I talked about you early in your career started by building data centers, and I'd be curious to know what that looked like, you know, like, uh, back in the day. Eric Lamb (00:10:32): I was one of the original group at DPR that had experience in building data centers in the 1980s. In fact, I was project manager for the, what they called the network operations consolidation project when Wells Fargo bought Crocker Bank in San Francisco. And they had a downtown, on Fifth Street, south of Market data center that was massive. It was like six floors high um, it was all raised floor, and it was all 3060 IBM mainframes, these massive computers all over the place. And in those days, they were so massive and they were generating so much heat, they used water cooling, liquid cooling to these mainframes, which of course then we transitioned over the subsequent 10 or 20 years to air-cooled data centers. And now with GPUs and the current generation of data centers and the heat, we're back to liquid cooling. It's funny how some things come back to the beginning. But we had a basement full of chillers generating the cooling lines to go to all of the mainframes. So we had, an entire floor of disk drives, you know, for the memory and for all of that. It was, was something else. But it is something, as we started the company, we started to do some data center work, although for DPR in the '90s, our advanced technology core market, which includes data centers, was doing more semiconductors. And in those days, semiconductors were still, being made here in the US. It wasn't until maybe 10 years later that most of the semiconductor, uh, manufacturing went offshore. But, we were doing work for a lot of major semiconductor companies, and that also allowed us to grow across the country because we started doing semiconductor work in Arizona, Texas, Virginia. We went up to Oregon and to the Northwest to do a semiconductor facility. And oftentimes when did those large projects, uh, people stayed behind and, and set up offices. The data center market has evolved quite a bit, you know, since even the 2000s when-- in early 2000s we were doing colo centers and small, uh, data centers of, uh, servers, to do this sort of edge computing with the colo centers. But now the, the scale of the data centers is much bigger, and the, the pace is dramatically faster, not only than other data center work in the past, but any other construction work we're involved in. So it is an area that a lot of contractors and builders are into because there's so much of it going on. Uh, despite some of the, community sentiments opposed to the data centers, uh, as sort of a proxy for people's opposition to what may become from AI. AI data centers are, are obviously getting some pushback in some areas. I think ultimately uh, they will continue to be built out. We see, uh, a good backlog for the industry going past 2030 with this, uh, AI infrastructure that's currently underway, and likely far beyond that. But it's going to be not without, uh, a lot of, issues in trying to solve them for the communities and for the country and what we need to do for power and compute and that sort of thing. These projects, these what I've observed is they just take a lot closer coordination because they're so fast between the design, the owners, the equipment suppliers. You know, you've got prefab elements, and everybody needs to be very well connected, and the trade partners are s- extremely critical to be on early with the general contractor and the design teams and all of that. So it is something that I think, is actually a good thing for the industry because of the level of collaboration and coordination that's going on between all these parties and the amount of design build that's going on, I am hoping that that will spread horizontally to other markets and other aspects of our industry because, uh, there are some great success stories happening across the industry with lots of builders and owners collaborating with these very fast projects. Kanav Hasija (00:15:06): So Eric, I had a follow-up on that data center thing, right? Which is when owners and contractors have a will to move faster, they do move faster, and data centers do get built out faster, right? So the sentiment of the industry that we have gone slower in constructing in the last two decades, is being reversed with the data center build-out because we can build things faster. So from the learnings of how we build things faster with data centers, are we coming in our own way in other build-outs as an industry? Like are we being our own bottleneck? Are we creating our own bottlenecks? What's happening? Like what's the learning from there to apply to commercial hospitals, industrial and so on, right? Eric Lamb (00:15:48): Well, certainly the adversarial nature of the business where, certain people are not trusting other people, uh, has been a bottleneck or, or an obstacle to delivering things as fast and, as complicated as some of these data centers. I mean, we built a 100MW data center in Texas for one of the NeoClouds, 100MW in 10 months to first, uh, you know, racks getting plugged in to the data center. But that would not have been possible without that entire group of, entities, owners, trade partners, designers, everybody working together from the beginning. And so I do think we have gotten in our way in this industry. I think you're right. If we can develop a process that is more sort of, uh, collaborative, trusting relationships, uh, I think we could accomplish so much Jon Wright (00:16:46): Switching gears, WND Ventures and Construction Tech Investing, Eric, you're, uh, playing a lead role here, in that DPR's effort to do that in that, company. WND has backed companies like Trunk Tools, Dusty Robotics, and MeltPlan. What's your filter, uh, these days for, construction tech and making bets? And I would just say, listening to your background, your whole career you've been forward-looking, building data centers in the '80s and early '90s. You've kind of been ahead of the curve. You've seen where, things are headed. What's your pattern recognition tell you these days in terms of making venture bets on construction tech? Eric Lamb (00:17:25): I've never been in our industry in, in a time where I've seen so many startups and so many different construction technology companies coming into our industry. So it is, uh, quite, incredible. I do think, the common answer you'll hear from investors like me will be always that there's some painful and recurring problem that's being solved by what the, tech company is involved in. I think with construction, they've gotta have a sense of what's happening in the field and not just the technology, and somehow the product has to sort of fit naturally into how projects are delivered right now. So you don't want to go into construction companies and, change a lot of what they're doing. Now, it may ultimately change over time, but you have to somehow fit into the workflows, uh, of what they're doing now, and even some of the current incumbent software that they're working with. So I think those are really important parts of any kind of new construction tech investment. I think the other thing that we've observed that's almost... it's somewhat nuanced and, and a little bit, anecdotal, is do you feel a pull? Do you feel, attraction from the construction people of, "Gee, I wanna use that. That could really be useful to me"? Jon Wright (00:18:47): You're getting a lot of construction tech pitches, uh, as you mentioned, and you've had some in the past. What's one that you've passed on that you still think about? Eric Lamb (00:18:56): This one's an easy one, and it goes way back to the beginning of, our investing in it. We, we started WND Ventures quite a while ago. Over ten years ago, we started, WND a-as a, as an entity, as a corporate venture capital for DPR. But we were investing actually prior to that, in a couple of companies. We had the opportunity with one company, in about 2012 that had a, uh, ability to put drawings on an iPad and people work in the field. So it was Tracy of PlanGrid, who came into the, boardroom at DPR, and we were already starting to get some users, who were pulling out their credit cards and using their iPads with PlanGrid for, you know, doing observations, and updating changes at night and all that sort of thing with, PlanGrid. It was the early days of it. But we, we saw their pitch in the office, and they only wanted like a hundred thousand or something like that. I don't remember what it was, but it would've been easy for us to invest. But there was a couple people with us that felt that is easily replicated by Autodesk or by one of the big software incumbents, and this will get, snuffed out within the next year or two. There'd be no point in investing because they'll just do that. So, the rest is history. Of course, they ended up selling the business to Autodesk years later, 2018, I think, to, for eight hundred and seventy-five million dollars. And then we ended up, uh, having PlanGrid as an enterprise agreement across the whole company. And then, of course, now we use it at, Autodesk as a part of the Autodesk suite of products as an enterprise. But, you know, this is one thing where you learn your lessons in the sense that feel that there was pull from the field that we ignored, and we did not... and I think there was two lessons. One is we didn't pay enough attention to the field and what they found useful. The other, the second thing is we put too much trust in a large incumbent software company being able to just do this themselves. We think there's a lot of what's happening right now with what MeltPlan is doing with a lot of other AI startups, the incumbents are not capable of doing. I've lived it now ten years and seen the other side of this, but that's one that I still remember Kanav Hasija (00:21:25): Tracy is great. She's a great founder. You're right, Eric. Um, and kudos to Autodesk, Revit and PlanGrid were two crazy good acquisitions. Like I still remember Revit was acquired by Au- when they were doing one million in revenue, Revit. Eric Lamb (00:21:41): Yeah. Kanav Hasija (00:21:41): And now Revit is the main design product for Autodesk PlanGrid was doing like I think, 30, f- 40 million in revenue, and that's the whole, uh, construction cloud now. So, so yeah, kudos to Autodesk guys. You're right. But that's, that's so true. I mean, every cycle we see that incumbents can't build the next technology, and then they acquire something down the road Jon Wright (00:22:03): I mean, if you want to, great, uh, principles there to think about, pull from the field, no-notably. What are the GCs actually needing that the market is not providing now? Eric Lamb (00:22:14): I think we're at the beginnings of the AI sort of cycle, right, for the industry. There are some good tools out there to do, design reviews. Uh, obviously you guys are doing bid leveling, contract reviews, other kinds of drawing analysis, sharing information, things like that. I think there's emerging companies that I would call that are starting to use agentic workflows into existing workflows to streamline processes inside, our companies. And I think that hopefully by the end of the year, beginning of next year, we're gonna see some really good progress in that area. I think the area that, that we lack right now, and I think if I separate construction from the front end, design and concepts up to, say, a control estimate the pre-construction, the design, the estimating, all that. Then I put the second part is everything in operations and in the field. I think the operations and I think there's some good companies working on pre-con like MeltPlan and others trying to figure out that whole process and helping support those workflows. What we lack right now are workflows for the field. There are companies out there with AI tools that are trying to help submittal reviews and some other, uh, procurement and that sort of thing. But there's not any agentic workflows that gives you the financial picture all the way through the course of the construction in the field. And so that's the area that probably is most lacking that I see right now. And it's also the hardest because you've got to integrate with ERPs, you've got to integrate with, whatever spreadsheets people are using and all those workflows that are going on during the course of construction. So that's the area that I'm very hopeful for next year and beyond of what could ultimately... There's a lot of time people spent on monthly projections, forecasts, labor productivity. The economic prize to me is the company that connect all those workflows that have to do with money and changes and all of that through the course of construction. That's what I'm hoping for as I look forward to AI in our industry Kanav Hasija (00:24:38): You know what, Jon? Eric said this and also, the DPR current CEO, George Pfeffer, know, they both told this to me. It sits in my mind forever now. It's like engraved. When I think of a general contractor, I think of thin margins. Like, they live on thin margins. They have, a 2 to 4% net margin. And when I ask them like, "How can you increase those margins for your business?" told me, that's engraved in me, is like, A construction project does not incur one big mistake. It incurs thousands of small mistakes. So the construction margin is made up of thousand of small mistakes avoided. Any technology that can, both in pre-construction and construction, that can, avoid those mistakes, thousands of those, not small, small mistakes, because there's so much data, the thousands of sheets and submittals and ten thousand documents, and one human cannot impossibly, you know, get all that together Jon Wright (00:25:36): TK, you know, putting your, uh, project exec, project manager hat on, sitting in the trailer out in the field, what comes to mind when Eric talks about, agentic workflows that could increase efficiency there? Let's take that from, like, the real pain to what, what could be. What comes to mind for you? Tanmaya Kala (00:25:55): One thing like ju-just kind of stepping back, as a project manager or whoever is leading the project, could be an executive, could be a senior manager, could be a project manager, it's almost like a CEO role because you're managing the finances, and there is a building being built. There's an engineering component to it. There's a building going up. But your tie back to corporate is the financial management, and a lot of that is how much you got paid and how much are you projecting to get paid and how much change work you did, right? We had this joke, like I was one time running a very complicated project and Eric would ask me, "What, what's, what do you think our margin is gonna be?" And I'd be like, "Well, it's this and that, and this issue happened, and that issue happened." And Eric went on the whiteboard and said, "It's simple. It's A minus B equals C. A is your revenue, B is your cost, C is your profit. Why can't you figure this out?" And I'm like, "But there are like million things going on." But end of the day, I kind of like spending all the time with Eric on that, I finally I understood that it is that, it's about how you figure that out efficiently. So every company has a process of every month the project lead provides that financial forecast. How good you are at it depends on how accurately can you project forward. But for that, you also need to know how much you're spending right now, what your burn is. A lot of those things are right now done manually, right? At least back in the day, we, we look at bunch of cost reports and, and we do some projection. It's a lot of spreadsheet-centric process, with bunch of human man-hours, and I do think that's a real pain that can be automated because a lot of that is how your invoices are flowing, where are you tracking them, are you tracking them appropriately? Just bringing it down to that A minus B equals C for the executive level. Jon Wright (00:27:49): I can tell you as the owner, developer, you know when those conversations have happened because then you get that email or call to like, "Hey, we need to talk," uh, which typically means some adjustment. And from an owner perspective, I think that having that information sooner, you're gonna reduce that risk of an overrun. Having that benefits you f- as a GC for your margins, but us as a project owner on our risk exposure on cost of like, I know they're gonna call us because I've been watching this happen, or we had a lot of rain, or, we couldn't get that, material delivery, and I know that slowed things down. So, I think that redounds to everybody. Uh, and if we can get that sooner in the cycle, the better, I think. Eric Lamb (00:28:38): The area that I think is u-underrepresented still in the AI startup realm is really around, labor. And when I say labor hours for trade partners and self-performing general contractors, tracking labor and also looking at labor productivity and tracking, even how to code the time every week is, is a recurring issue that people either don't code it properly or don't code it at all. How those labor hours can be associated with different safety data. What are the near misses? What are the observations? And understand how labor and safety and labor and productivity is all tracked. There's so many things that could be automated in the tracking, and then using your past data to learn in order to help you with like, uh, TK, you're saying about forecasting and understanding, identifying risk early on where your labor's going, which is a big deal Tanmaya Kala (00:29:39): When I say it's pain right now, I still debate how much AI can help other than providing good data for someone to do a projection, because projection ends up being a very human endeavor, and I've seen both sides of it. Optimistic people can be too optimistic. Pessimistic people can be too pessimistic. It all depends on, you know, what projections you make and whether they pan out or not. Kanav Hasija (00:30:04): I had this question for Eric, I'm writing a piece right now, Eric, for tomorrow's update. But there is a boardroom and operator divide I'm seeing in the industry, not just construction, everywhere. Where the boardrooms are saying, " Let's push AI down to the last level. Let's use AI as much as possible." At the operator level, I'm seeing two different philosophies at the same time. One, the operator is fearing that their jobs will go away with AI down the road. And second, at the same time, I'm hearing from the same operators, "AI is so sloppy. It makes mistakes. It hallucinates. It's not there yet," right? Is that boardroom and operator tension being managed right now? And then the operator conflict of AI will take my job away versus AI is sloppy. Like, how should, uh, an operator and an executive, their path going forward in, in this world of AI? Eric Lamb (00:31:00): I think we're in the early stages where people need a few wins to show some efficiency gained in a particular process or two. And that may be some single point solutions that are out there in AI, whether it's in design reviews or other things that are helpful, that are saving a lot of time and hours in being able to demonstrate that. Those are not ultimately the type of agentic workflows that work autonomously inside of your systems to help you and assist you in forecasting and, other sorts of things we just talked about. Those are emerging still, and I'm excited about where those can go. I do agree that there's been, uh, in some of these areas where we're attaching to different systems and data it's very complicated, like we're all saying. And even on a total cost of A minus B equals C, you're pulling in the job numbers, but wait a minute, the agent didn't pull in the sub-job numbers also, because those are also a part of your total revenue, your A, then you've got the wrong numbers, the errors you're talking about. So it requires this almost forward deployed engineer of AI sitting next to the field person and sitting next to how they do it and what they do in order to get it all pulled in correctly and then built. Because if some company off on the side is doing this by themself, they don't really understand all the different sources of data. Like TK's talking about, some data at, at DPR and a lot of-- builders are spreadsheets. Our resource, projections on all the overhead, uh, general conditions are spreadsheets. And so you've got to pull that in as well as things from the ERP. It's just a very difficult problem to work through, but it requires almost this forward deployed engineer sitting next to you because each company has a slightly different, data sort of structure and tech stack that needs to be worked through to get those agents correctly reporting out Kanav Hasija (00:33:06): It's the short-term pain for the long-term, benefit, which is AI is not a magic wand. You have to fix it with the right data and the right context, and that fixing of AI to your context takes effort, which you have to put in the short term. You have to get outside of your job fix AI for you in short term to get the long-term benefits, right? But if we expect AI as a magic wand and it's gonna do magic tomorrow, like that's not the answer. Tanmaya Kala (00:33:34): O-one question, Eric, I had was, going back to, uh, back to the days of founding of DPR. You were leading the preconstruction side and, you were responsible for creating some of the early processes. I would just want to ask, what were your goals for those processes back in the day? Like, what were the things you were trying to solve for when you created those? Eric Lamb (00:33:56): It's interesting. I, I, I think I had mentioned earlier that we were trying to go digital as a company in 1990 with what was possible that was digital in 1990. We did, the previous company, which was on spreadsheets, and before that was on just written, spreadsheets, and went into, uh, using an estimating, software. In this case, it was Timberline Estimating. And we developed, instead of using a standard means database, we wanted to make sure we had our own database based on our own data and performance as a company. So this was 1991, and we hired a guy to work full-time basically capturing data from our operations in the field on drywall we were doing or cost of steel or whatever. And we developed all of our own database. So we made a big investment early on. From our standpoint, that was a big investment, uh, of putting somebody full-time on building that database from which it became a shared database that we modified for different geographies as we grew the geographies. But we did that for a good 15 years and continued to update that database. So for us, we wanted to be able to learn from the field from what the data was telling us to get back into a database, uh, a common database format. That was very successful for us. I think in the '90s and early 2000s, we were considered quite, innovative in our ability to do these, uh, estimates that learned from our operations, uh, performance in the field. And that's an area that I think with AI, you can do that even so much more effectively by automating the capture of that data, scrubbing of that data, and understanding it, and then be able to move it forward. Because we've never been able to capture everything and all the issues in the manual way that we were doing before. But I'm really excited about the future of pre-construction and companies like MeltPlan and where you can go with AI in not having to reinvent the estimate every single project and not learning from the, you know, data in the field Tanmaya Kala (00:36:15): Or relying on what we remember we did in the last project. Kanav Hasija (00:36:18): TK, you have been working with Eric, or at least at DPR for the past 13 years, and you had a chance to work with Eric. Tell us one anecdote which never goes away from your mind. Tanmaya Kala (00:36:31): One anecdote that comes to my mind is, um, I was, uh, I was, I was leading a project, it was a pretty complicated project, and I got a call from Eric. Eric was president of the company at the time and said, "Hey, I wanna know your projections. What are the revenue projections? What are the profit projections?" Now, here's the thing. I talk about being a project manager is like being like a CEO, but also when the president of the company calls you, it's time to start making your resume sometimes. Like, especially when they're saying, "What are your financial projection?" I was super nervous. I showed up to the Redwood City office with our finance guy, with this whole complex presentation of like all the things that are going on and all these issues, like the death by a thousand cuts we talked about. And Eric just shrugs his shoulder and s-says, Eric just shrugs his shoulders and says; and Eric's a very direct and a blunt guy. He's like, "What the f*ck are you talking about? It's simple. It's A minus B equals C. A is your revenue, B is your cost, and C is your profit. Why can't you get it?" And I'm like, "Uh, well, it's, uh, it's a little bit more complex." He's like, "No, you have a week. Figure that out and then come back." Like, all right. So, um, anyways, that I remember, and it really forced me to think through and simplify things and, that one week of learning was like getting a degree, so, management. Yeah. Jon Wright (00:37:55): Sometimes your excuse is, "I didn't have time to make this simpler." Kanav Hasija (00:37:58): Yeah, I mean, being blunt is the most useful thing a mentor can do to their mentee, because there's so much learning in there, right? So I think the last question we have for you, before we let you go is for the executives at a mid-sized GC firm or a growing high-growth GC firm, what's an, an advice or tool you have for the executives at a general productive firm to grow faster? Eric Lamb (00:38:24): So th-this is an area that, is so important because what I've learned over the years is, you can't take things on, uh, personally, meaning, you need to work as a team, and you need to find people to bounce things off of and really trying to help, solve problems and look at projects and figure out, how to go forward. So that's the area that I think general contractors need to make sure that they are able to do and how they're structured and what they're doing. I think the other thing is having a culture that really rewards the quality of people's questions versus everybody having speedy answers. And so I think those are areas that I, I really try to look back on my career and talking to other companies and contractors that sort of the value of this teamwork and having the ability to ask questions and, understand the future, I think those are the areas that I would value the most Kanav Hasija (00:39:28): Awesome. There's so much wisdom in there for the last, uh, 40 years, Eric, these are the kind of things that matter more than technology or data is the, the teamwork and the core values. Eric Lamb (00:39:39): I think the thing we were talking about before that I'm really excited about what you guys are working on sort of this, problem in our industry is this failure to capture and reuse lessons from completed projects. We tried to do that in a knowledge database back in the '90s. We were trying to do something. Atul and I worked on it. It's probably late '90s, but we never really quite got it figured out to where people could share it. I think if you could, like you were saying, all the death of a thousand cuts that happen on all these projects, if you could find a way sort of project becomes a learning system that then you're putting back into it that everybody can reuse, I think is gonna be awesome. Kanav Hasija (00:40:23): Yeah, that's exactly why I got excited to join this industry is because, no, wrongdoings of the incumbent technology players, but before AI, the technology could not understand documents and images, this industry is full of that. Now, AI can go deeper and understand those things and parse things out. It's a golden opportunity for this industry Eric Lamb (00:40:47): I would agree. It's the perfect industry for AI when you think about it, because of the complexity and the number of issues and fragmentation of different companies. Kanav Hasija (00:40:56): Awesome. Jon Wright (00:40:57): All right. Thank you Tanmaya Kala (00:40:58): Thank you, Eric Eric Lamb (00:40:59): Alright guys