Rishi (00:41)
Today our guest is Tanmaya Kala TK. TK, welcome.

Tanmaya (TK) (00:48)
Thank you, Rishi. Thanks for having me on the podcast.

Rishi (00:51)
First section is on your background and journey into construction. You spent over a decade at DPR construction and eventually became a project executive managing billion dollar portfolios. What first pulled you into construction and what made you stay in such a demanding industry?

Tanmaya (TK) (01:13)
Ooh. it's a it's a great question. So my background, my undergraduate was in civil engineering. And j just for context, my grandfather was a civil engineer too, and and he built large hydroelectric projects. and growing up, there was a lot of influence from him where we would drive around and my grandmother and my parents would be like, your grandfather built that and your grandfather built that.

As a kid you thought he single handedly built it, but as you spend more time in the industry, you realize it takes a village. so so that I always found that attractive. And so after my civil engineering, I actually had an offer to join a software company. but I decided to stick with civil engineering and it's it's been a very rewarding profession.

Rishi (01:44)
Mm-hmm.

Tanmaya (TK) (01:58)
I actually like the second part of your question. What made you stick around with this demanding industry? Even though now I have a startup, so I have gone away from it a little bit. But again, we're still building tools for the construction industry. I think one thing I would say when you're in construction, you always have a feeling of like you're building something real. You see something physical going in front of you that you can touch, feel, and walk in. and and it's just a very different feeling.

And second thing I would say it's a very people oriented industry. So if you enjoy talking to people, interacting with people, construction industry is just an awesome place because it's about working with people to build this physical thing together. And you're together in one place doing that. So that that it's just an exciting thing. And I I I would say this for last 18 years I was in the industry, every day I learned something new.

Rishi (02:52)
Right before

this recording I was sitting with a builder. The conversation is so different. he's managing this foundation and trying to make sure these complex entities they come together.

Tanmaya (TK) (02:55)
Mm-hmm.

Mm-hmm. Yeah. No, it it it i it is it it's like it takes. I mean, I I don't know how much details you want me to get into, but there are 40 entities that usually are involved in a construction project. Forty entities as in 40 different companies. It could be 40 to 50 depending on your project size.

when you're running a startup or you're in a corporation, you have people problems because you have different people working together, but at least the company is the same for the most part. Here you have forty companies and each company has different people. So it it just exponentially is a different ball game. It's it's hard to imagine once you're in it.

And all these companies are tied together with these subcontracts and and like you're aware of like, you know, the financial side is all tied through these contracts. So these are like thin threads tying these things together and how do you do the dance is what makes it fun.

Rishi (04:01)
Right. You've worked across structural engineering, project management, and executive leadership. How did your perspective on construction change as you moved from the field into leadership roles?

Tanmaya (TK) (04:01)
Yeah.

Yeah, so I'll break this question into two parts. So one is structural engineering to construction. So when I was a structural engineer, my deliverable is a set of drawings. So so in some ways you're a degree removed from the final product, which is the building itself. You're you're more designing the building on a on a on a CAD and then on a piece of paper. the construction guys have to put all of that together.

So a little bit of a different mindset. Like I'll I'll be honest, like as a structural engineer, I have been guilty of the control C, control V problem of copy pasting details from other projects that kind of looked similar and then kind of reviewing it and putting it in rather than drawing it from scratch, right? Which construction guys complain about. And I've been that construction guy and when I joined the construction side complaining about

This detail is not for this project. This detail doesn't even match the project. What the hell? What is this guy doing? When I know what they're doing, because I've I've been guilty of doing that too. construction from a field to a leadership role. one thing I say, in some ways it's very similar, in some ways it's very different. I'll tell you where it's similar. In a in a construction industry, if you start on a general contractor side, you start as a project engineer.

Rishi (05:07)
Mm-hmm.

Tanmaya (TK) (05:30)
You have to have a certain level of leadership day one because you're still managing work being put in place. you're never not managing. You're never behind a keyboard and and just not aligned with what's going on in the field. So there is always a degree of management that's needed, which stays with you throughout your life.

In a leadership role, I always go back to this book. It's one of my favorite books, Good to Great. And at DPR, they used to really motivate us to read that book. And it's a great reading, which talks about five levels of leadership, which is really important construction. One is being a competent individual, which is what a lot of field is. Like you need to be competent at your job, being able to write RFIs, solve problems because problems keep coming. You need to run out there, solve them, communicate them well. Second is being a good team member.

Again, you need to have your team members back. You need to be able to know when to jump in, when to hold back. third is you need to be a good team leader, right? Which is how do you motivate people around you? That that's more of a project manager level, right? And then project executive is what I would call you have to be a good manager of managers. Because managers think a little bit different because

to rate how good a manager is doing is a little harder because you have to rate them on what the team around that person is delivering, versus rating an engineer is a little easier because you're looking at what they are producing. Right? That's that's more measurable thing, if I can call it that. So being a good manager of managers is a it's a different different way of thinking where how is this guy empowering his team at the same time.

Rishi (07:01)
yeah.

Tanmaya (TK) (07:14)
not being just a delegator, right? It's a balance. And and how do you how do you do that effectively is important. So that's the people side. And then there is the project side. And project side, honestly, when you become like at an executive level, you're more focused on making sure are am I delivering value to my client? Is my team set up to deliver that value to the client?

Because as a project manager, you're more focused on building the building itself in the most efficient way possible, which we have to talk to the client what they are looking for, right? They they might wanna add an MRI machine, middle of the project. Well, that's the most inefficient thing to do. But long term, that might be a great decision for them because they need to have this newest, latest MRI because it's gonna generate a lot of revenue. then

a engineer is more solving day to day problems. They're really focused on their trades that they're managing, the scopes they're managing. So your field of vision expands as you keep growing in the organization.

Rishi (08:18)
Yeah, the client has somewhat different objective compared to let's say G C.

Tanmaya (TK) (08:22)
Mm-hmm. Yeah. Yeah.

Rishi (08:26)
Many people see construction as just building things, but the reality is much more complex. What did you learn about the industry once you started managing projects end to end, from selling the job to delivering it?

Tanmaya (TK) (08:40)
Yeah, this might be a little controversial depending on who who's listening, but I think construction is probably one of the most complex endeavors we take as human beings. And I know, there are people building AI models and people doing a lot more things that you know are considered as very smart and very intellectual. But

When you're building a building, there are a lot of moving pieces and you're touching a lot of things. it's all this material need to show up at a certain time, and all these people who are gonna skilled people who are gonna install that material need to show up at that time too to put this building together, which has millions of pieces in it. So you're managing material supply chains, you're figuring out your people.

there are external things like politics associated with a project. Like someone might not want that building there. you know, like building a building as any of us have seen someone working on the road is is for the community your building is in, is always a bit of an headache. Right. Ca you know, when the whenever there's construction going on around you, you're like, what are these guys doing? What's going on? They they blocked the lane, you know, there's some dust flying around.

There are all all these things that you have to manage through as well. And in some ways, I feel like the technology side is also getting exposure to that as these data center builds become more and more complicated. Like I'm hearing podcasts where you have people like, even Sam Altman and these guys going like, Yeah, building a data center, it takes a lot to figure out how to get zoning right, the permitting right, to build a building on time, get all the material on time. So people are also understanding.

Yeah, in a way the complexity involved with building these buildings.

Rishi (10:26)
Yeah, actually I just was building a robot the other day. And I'm a software engineer coding and deploying and all. And once I started building that robot, you know, putting together these mechanical, electrical and wiring and a battery together, it gave me a different appreciation for building things.

Tanmaya (TK) (10:44)
Hm. Yeah. Yeah. Yeah. it's not the easiest thing to work with your hands.

Rishi (10:48)
The next thing

yeah, it's tough. And a lot sometimes stuff didn't work. You think that you plug in this wire and this battery and suddenly things should work. Sometimes stuff doesn't work and I don't know if the wire is not in right or like the battery is faulty, who knows?

Tanmaya (TK) (11:05)
Yeah.

Rishi (11:05)
The next section Tanmaya here is on the finance versus operations gap. One theme of this podcast is the disconnect between the job site and the finance department. From your experience running large projects, where does that disconnect usually show up?

Tanmaya (TK) (11:23)
Yeah, I mean you we started this podcast about you asking me what attracted me or what makes me stay in the construction industry. One thing I can definitely say it's not the finance part. Like nobody becomes civil engineer or comes to construction when you know, if you ask a student who's studying civil engineering, he's not going like, you know what I'm excited about? Doing invoices. You know what I'm about? Doing billings, right?

it's not the sexy part of construction. it's not the part that that people really signed up for. Having said that, it is a very necessary part of construction. I I always used to say, like, what makes a good project manager a good project manager is who knows where actual cost is at compared to the budget. That like if there is one thing that makes a project manager strong is knowing that. Because end of the day

Everything's a business. We're trying to make money. We're not just, doing things for fun and games. And and and so it is necessary part of it. But having said that, there is something in us, and it's it's probably true for all human beings, if there are things we don't like or we're not interested in, there is a tendency to procrastinate, there is a tendency to deprioritize it, and there are excuses you can make.

And in construction, it's very easy to do that because every project in some ways is managed chaos. Project manager every day is dealing with multitude of problems and it's all about prioritization. What am I gonna do now versus what I'm gonna what's gonna take a backseat to solve this problem? And there will always be a a psychological leaning towards solving a problem that's out in the field, because that's what

you know, we as construction managers and civil engineers feel like is the real thing and that what we are paid for. And that could lead to deprioritizing what if an accountant and saying, Hey, you didn't review the billing, you didn't code the invoices, you know, I need all the coding I need the all these invoices match to the cost code and you have to read the invoice. It's just those things would would start taking a backseat. And that just

That's the nature of the beast. And more disciplined project managers are better at it. That's how they grow. but you know, not always the case.

Rishi (13:41)
Project managers often focus on building the project, while finance teams focus on cost tracking and billing. Why do you think these two groups often struggle to communicate effectively?

Tanmaya (TK) (13:54)
Well, I mean before I answer that, I would say all good projects have a good streamlined communication with project management team and accountant team, accounting team. I think it's it's very important. And at least in my past that's where I've seen a lot of success. the communication breakdown when it does happen.

and it happens on many projects, by the way. I shouldn't say, I mean, I'm talking about successful project versus most projects. It's just it's different priorities. Like the the accounting team to them, and so I'm gonna kind of there are there are three three teams depending on the side. If you're on a large project, there's an accounting team that's focused on billing and making sure that you know people are getting paid, like you know, owners play paying the GC, the GC's

Paying the subcontractor on time, the lean waivers are getting done and all that good stuff. Then there is a cost control team. There might not be a cost control team that might fall on project manager, where they're making sure all the invoices are being coded properly and all that good stuff, like all the interim stuff, like cost control itself. the the breakdown to me, it's I think the simple answer is it's just different priorities. Yeah.

Rishi (15:05)
Yeah. Also these people tend to use different

systems that do not talk to each other. Project manager is in his silo and the accountant is doing it in his silo and the two things, you know, they're not talking to each other.

Tanmaya (TK) (15:19)
Yeah. No and and that this is something industry has been trying to

solve for for some time. I mean, there are tools that try to merge the accounting side and the PM side as much as possible. I mean it's not merged, but at least being able to pull reports, like at least being able to consume the data from the accounting side. Accounting side can do a lot more on the PM side, to be honest. because we trust them more with the money side. but

Yeah, no that that that's true. Like I I mean even when I was an executive, I don't think I ever had direct access to accounting. I can generate reports from it and dashboards from it, but not edit anything. I had no edit right through it. Yeah.

Rishi (15:59)
Yeah. You said something

really valuable. you know, the companies like Procore, they're huge, they're project management companies, but they're still not system of records in these construction companies. We talk to construction companies all the time. There's the project management stuff, okay, they're doing something on there. But really the source of truth is the accounting system.

Tanmaya (TK) (16:20)
Yeah. And that's true because PM can be different because you're also projecting, right? yeah, I'll get paid for this. yeah, yeah, that I know a change or is gonna get approved. You know, and and and usually the and this is a good thing, but can be a bad thing too. Project managers, successful ones are generally optimistic people as well. because that's how they became successful.

Rishi (16:27)
Mm.

Mm-hmm.

Tanmaya (TK) (16:44)
and so there is a tendency to be like, yeah, yeah, yeah. And whereas accountants are like real, like like, did we get paid for it? No. well, you're not like we have it, like this is this is the real this is your real financial position, by the way. Right. but I I call I call the client. I know, he'll pay me. I I got I I got it, you know, but the the account you know, so so there is there is actually talking about difference in how people operate, there's that.

Rishi (16:55)
Mm-hmm.

Exactly.

Tanmaya (TK) (17:11)
but yeah, I mean having different systems is is probably part of it, but it's also mm, I think different mindsets to how how these two cohorts operate. Yeah.

Rishi (17:22)
Yeah,

I project managers, their job is to project too, right? The future. They're not operating on the history, right? They're not keeping track of everything that has happened, right? They're thinking, How could I make this project successful? It sometimes requires looking back to the history, but it also requires looking at the future.

Tanmaya (TK) (17:26)
Yeah. Yeah. Yep.

Yeah. Yeah. I mean, one of the main jobs for a project manager is to look to the future and and see but I mean, the thing is it needs to be grounded in reality. That's where I said it's g it's good to have access to accounting side as a read because you can see, okay, this is where I'm at, this is how I'm gonna make the project whole. Right. Maybe I can't make the project whole. Then I need to tell someone and you know, and and then, deal with the consequences of that. So

Rishi (18:02)
Yeah.

By the way, have you fired any project managers in your tenure?

Tanmaya (TK) (18:14)
I am lucky enough to not have done that. and by the way, I I've also worked for companies like McCarthy and DPR who tend to hire, best in the business. So that that helps. And but there are performance review mechanisms where you have to have these conversations where you might put someone on what was what's called a PIP performance improvement plan, which is more around like these are your performance.

Deficiencies, we need to see improvement here, and you give them a timeline to do to to showcase that improvement. So I've I've been through that, yeah.

Rishi (18:46)
Actually Actually

I was in a performance improvement plan, like in one of my jobs. It was basically a firing plan. It was not like a performance improvement plan. Listen, you got six months you have to get out.

Tanmaya (TK) (18:56)
I'll two people who went

through that with me, still with the company and they ended up doing very well. So I'm I'm proud of that. Yeah.

Rishi (19:06)
you mentioned that many operations teams see accounting tasks like coding invoices or preparing billings as a chore. Why does that mindset exist and what impact does it have on projects?

Tanmaya (TK) (19:21)
I mean they look at it as a chore because it is a chore. I don't know how else to say. It's like yeah, I mean I don't know how else to say it. but I mean, like I said, it's not the exciting part of construction. And and and so so what impact it has is it the the the invoices do add up a lot and then

you might lose track of like how much money you're burning because end of the day knowing your burn is extremely extremely important and and also one thing accounting actually does well is unless you code it right at least from what I've seen they won't pay the vendor and then the vendor will get your case. Vendor will be like hold on a second I'm not getting you more stuff like you haven't paid me and it's like why didn't we pay him or you never coded it. So

Rishi (19:55)
Yeah, yeah.

Tanmaya (TK) (20:05)
So there is a feedback loop it's not a great feedback loop. I mean it's it's a it's a horrible feedback loop. But what it does is as a project manager you figure out that hey, if you wanna keep your vendors happy, you gotta pay them in time, you gotta p to pay them in time, you gotta get invoices coded in time. And that's how it

Rishi (20:11)
Mm-hmm.

Yeah, and product managers can code too, whatever. You know, you just got like you know concrete delivered and you can say that it is like excavation stuff.

Tanmaya (TK) (20:33)
Yeah.

I mean you know, that does happen. It's not it's not great, but that can happen. again, a good project so we so like I mean, like you're aware, like there are cost codes, there are billing codes. we need to be tracking cost to the right cost code for us to know where we are overrunning our budget, right? Or our estimate.

Rishi (20:54)
Mm-hmm.

Tanmaya (TK) (20:57)
And and where I might have savings. So so actually speaking of impact on the job side, what you're saying actually that that reminds me that does happen. And when that happens, it can land you in a lot of trouble because you might go, I have so much savings in concrete. Well, because we didn't code it in the right place. But you might have miscoded.

Right. You you might have put something related to structural steel into rebar or something like that. You might have made that kind of mistake. So it's very critical to code the right thing to the right cost code to run a project effectively. Yeah. Yeah.

Rishi (21:32)
Thank you for saying that.

the next section here, TK, is real job site finance stories. When you were managing projects, you mentioned spending entire Fridays coding invoices in the project management system. What did that process actually look like week to week?

Tanmaya (TK) (21:41)
Mm-hmm.

Yeah, so actually this this this kind of ties back to the last question you asked, right? Since it is a chore, best way I at least like I mean this is a personal tidbit, best way I have found to make sure the chore gets done is I would block my calendar for doing as a project manager doing all the invoice coding and and I would pick a Friday. And usually usually Friday afternoons are a quieter time.

Rishi (22:01)
Mm-hmm.

Tanmaya (TK) (22:21)
for on a construction project. So I would block that time out that I have like this is when I'm sitting in front of my computer and I'm getting this done. Right. And it's like it's like butts on seat thing. Like, you know, I'm gonna sit and get it done and until I'm done I'm not gonna leave. So that helped me get it done, keep my accountant happy, help me get good

cost reports on what the burn is and where it's getting coded and all that good stuff. Because sometimes you're coding an invoice, sometimes you figure out you need to add more codes. You might not have enough cost codes. Like you like because the routinely you might find that this is not something I had anticipated or this is something I want to track as a separate cost code because this might be related to a change order that I might have to show owner the invoice because it was T&M work

Rishi (22:54)
Mm.

Tanmaya (TK) (23:14)
So data in itself can be very revealing if you force yourself to look at it. so when you're sitting down and doing it, it it reveals that kind of information too that you might not have anticipated before. So then when you start the next week, you're like, I need to talk to the client about this because these things were not anticipated, right?

Rishi (23:36)
There could be some value in doing the chores.

Tanmaya (TK) (23:39)
Yeah, yeah, exactly.

Rishi (23:41)
You

also described situations where accounting teams had to chase project managers to finish billings. Why does billing often become a last minute scramble on construction projects?

Tanmaya (TK) (23:53)
Yeah, that's a great question. th there are like a couple of components to it, maybe three components to it. So one is

Billing wants to be, at least the billing I'm used to, is you want to bill till the end of the month. Depending on your prime contract, and a typic typical contracts allow you to bill the work you have completed till the day you're billing, and you project out for the rest of the month. So usually 20th is the cutoff. So you're going first 20 days, you know what you've done, last 10 days you're projecting.

Obviously, there is a tendency for every trade partner to wait till the billing deadline so that they have the most accurate data, right? Or or they have the most knowledge of what they can get done because they want to maximize billing, right? It helps them with cash flow. So that's the trade partners thing. So they will wait till till they have the most data to to meet the billing deadline.

Then on the GC side, when we are reviewing, we have to stand behind, yeah, yeah, this billing projection is accurate. So it's just time sensitive because so owner has a deadline, we give trade partners a deadline, and there is a window in between where the general contractor has to approve the billing. all the trade partner billing, their own own billing, send it to the owner. They might get comments back from the owner, right? So

So all that cycle needs to happen in that little window, which is usually a week. And there could be many things going on in that week where which might create a situation where GC is not able to approve the billing. most likely candidate usually ends up being there is some change order that's about to be reviewed, but the work was done a couple of months ago.

So trade partner or subcontractor wants to wait till that change order is approved, so they'll delay their billing. Or you wanna make sure trade partner is whole. So they might have submitted a billing, they didn't have the change order, you're trying to get it approved, so you're dealing. So a lot of this confusion does happen because of these things. but general tendency is everyone's waiting till the last minute because they want to grab as much money as they can, especially for the work that's been done.

Rishi (26:07)
Cash flow is a big problem mostly w for subcontractors 'cause these guys they have to hash out this material and vendor bills before they can get paid by the G Cs.

Tanmaya (TK) (26:19)
Yeah. I mean and and it kinda goes back to the owner, right? So so GCs typically will have a net thirty with the with their clients. So so if I bill on June twenty-sixth, thirty days later, if my billing is approved on let's just take June thirtieth. If my billing is approved on thirtieth, I'll get paid July thirtieth. Trade partner has to submit billing to me by twentieth for me to make the thirtieth date happen.

For when trade partners submitting to me 20th, yes, they're projecting 10 days ahead, but they have already spent the money starting June 1st. So they are not seeing a check. So they're money out of pocket because they can't tell the laborers, hey, I'll pay you when I get paid. No, they have to pay, they have to pay their guys every week, right? You and I, like if we are when we have been salary's employees.

Rishi (26:57)
Mm-hmm.

Mm-hmm.

Tanmaya (TK) (27:12)
your business can't say, I'll pay you when I get paid. No, no, no, I'm an employee, you gotta pay me. Right. So labor has to be paid right away. Most vendors require you to make a payment within a week or two, right? Anyone, most material suppliers, right? So material suppliers, equipment rental companies, you have to pay them either right away or in a week or two. So you're a lot of cash out of pocket, and then you see the money 30 to 60 days later.

Rishi (27:37)
Thank you for adding that labor component.

Tanmaya (TK) (27:39)
Yeah. Yeah.

Rishi (27:40)
One of the more interesting stories you shared involved a subcontractor accidentally sending a certified payroll report with social security numbers exposed. What does that incident reveal about the risks in construction financial workflows?

Tanmaya (TK) (27:57)
that's an example of a human error, right? And since we are managing so many people, we also inherently have or our accounting teams, subcontractors mostly have a lot of labor payroll data, right? And it's a way it's very sensitive data.

And I think it's very important for them to make sure that they have systems in place that that data is masked properly, right? So there might be, you know, TNM jobs or cost plus jobs where you have to, for auditing purposes, reveal exactly how much you paid your employees, which is okay to reveal, but you can't reveal their.

personal information, right? So it's it's a balance, right? So I don't know what the answer to this query is. Maybe that's something for you to figure out, Rishi, but but I I think there needs to be better systems or or tools that that allow you to do this with more of a click of a button than relying on a human being doing it. Right.

Rishi (29:02)
Yeah, yeah.

And this a lot of these subcontractors, they're just mom and pop shops. They're like a husband and wife running like let's say five laborers. They don't understand IT, they don't understand so much about all the security stuff.

Tanmaya (TK) (29:10)
Yeah.

Yeah. No, and I think the this is a while ago, but I think it was a scan. We had to delete it from our system. We have to notify them and like to take care of it. But they clearly had it in like in hard copy somewhere and they just scanned it and saying, like, this is what I spent, right? But, you know, I mean for them to mask it they probably would have had to do it with a sharpie or something. Right.

Rishi (29:36)
Then it does.

They're like trusting you with all this PII.

Tanmaya (TK) (29:41)
So I don't

think they even thought about it, or like someone did it on their end. but anyways.

Rishi (29:48)
Yeah.

So did you guys actually use that PII for some malicious purposes?

Tanmaya (TK) (29:53)
Absolutely no. It was all deleted, we notified, everything was, you know. we had to tell the trade partner, hey, you need to watch out for these things. they didn't mention, we've done this. no, thanks for pointing out 'cause we've done this before with other people. No one said anything. I'm like, What?

Rishi (30:10)
Yeah.

The next section here, TK, is cost tracking and complexity. You ran cost plus projects where the client required actual cost documentation down to the square foot of a drywall installed. How does that kind of transparency change the financial workload of a project team?

Tanmaya (TK) (30:33)
yeah. So on a cost plus project, I mean I would say from a financial workload point of view, I would say lump sum is on one end, which I think is most efficient for financial side. And then cost plus is on the most complex end because cost plus projects require you to reveal your actual cost. and especially and also the I mean so it it can depend if it's just the GCs on actual cost or if it's

trade partners as well. In this case, trade partners were on Cost Plus too. that the financial workload goes exponentially up, right? On all sides. Like by all sides, I mean on the client side, on the GC side, on the trade partner side, or at least the trade partner who is cost plus. Because there are a lot of auditors involved, because yes, you're getting all that data, but someone has to process that data too. Right. Someone has to absorb it, understand it, and audit it, right, as well.

Versus so I I don't know d do your listeners understand cost plus lump sum pretty well or do you want me to quickly explain that? Yeah. So so lump sum is you hired me, I'm gonna say Rishi, I'm gonna build your house for eight hundred K and

Rishi (31:34)
You can talk about it. Yeah.

Tanmaya (TK) (31:44)
Every month I'm gonna bill percent complete, and and the beauty is in the eye of the beholder there, where what percent complete I think and you think would be different, but we can figure that out. The point is every month I say, okay, I'm 10% done. You might say, No, you're not 10% done, you're 8% done. And I'm like, Well, actually I'm 12, and I'm being, you know, it's whatever. But we are doing percent complete. You're not seeing my books. You you have said you're gonna pay me 800k, I'm billing you 10% of 800k, 80k this month.

I have a 10-month schedule that works out, right? And and and so on and so forth, right? Cost plus is you show me how much you spent on labor. You show me how much you spent on material. I'm gonna give you a fixed fee. You show me everything your actual cost. I look at your actual cost, I audit your actual cost, and then you get fixed fee on top. That's what I pay you.

So that is a very complex procedure because now, say it's a $500 million project, which has a lot more moving parts and pieces. You need a huge accounting team on your side to make sure trade partners' billing is accurate. Then client needs a lot of financial horsepower on their side to make sure what we are giving is accurate, all the paperwork behind it. Because you have to prove it, it's not just me saying.

Right. so so that's where it's an o it's orders of magnitude much more complex when it's a lump sum versus a cost plus contract. Yeah.

Rishi (33:15)
Yeah, very well expand explained

on the number of accounting people you need need depending upon what type of projects you're doing.

Tanmaya (TK) (33:22)
Yeah. Yeah.

Rishi (33:24)
On large projects, dozens of subcontractors are submitting invoices and labor reports. From your perspective, what are the biggest challenges in keeping cost tracking accurate in real time?

Tanmaya (TK) (33:37)
Yeah, so I mean like like I said, right? So mo trade partners typically in this day and age are lump sum. So so they fall in the person complete category for the most part. Again, cost plus we have already explained, so we won't go through that. So on lump sum contracts, when you're dealing with person complete, it can also get complicated on the size and complexity of the job because you have to go in there, someone has to

say this is what percent complete is owner would like to know or the client would like to know well what is the basis of when you say it's this complete how did you come up with that so there is there is that you have to figure out so there are companies that do progress tracking like you know doxel which will automate that process and make it more transparent.

traditionally it's based on a human being walking around the site and making that determination. where the other complexity is, but there could be a lot of TNM style change orders, time and material, where you do have to track time and material, get the change order approved. So that's where you have to actually show all the data behind it. so that's where it gets complex on a lump sum project as well.

Rishi (34:53)
You also mentioned material tracking as an quote unquote art and science. Why is tracking materials, especially across multiple projects and warehouses, so difficult for contractors?

Tanmaya (TK) (35:06)
Yeah. I mean, I'll start by saying I'm not as much of an expert in it as a trade partner would be, because they are purchasing a lot more material. But at a very conceptual level, like I was saying earlier, right, you're getting all these millions of parts and pieces that you're maybe let me back up a little bit. I'm gonna digress a little bit and then I'll answer your question. So there is a tendency

Of lot of clients and people outside looking into the industry to compare construction to the car industry. Right. Well, there is a you know, when I buy a car, I buy a car and I know the price and and and and and all that, right? Whereas construction things can things are always over all over like the timelines can change, the cost can change. In fact, there is this great book, how how big things get built. And

And in that book he talks about he he calls it the iron law of project management. He he did a research on 17,000 projects and found out that only two percent of projects hit cost time and benefits. Benefits meaning like you might hit cost, but benefits meaning client has to let go of something. They have to reduce the requirement to make you hit the cost, right? So

So 98% of the projects are not able to meet that. A lot of it is the uncertainty of a project in general. So, so now let's talk about material. In this globalized economy, it's not like in the old days where all the material is coming within three to four hundred miles of where you're building the project. Material might be coming from China, might be coming from Japan, might be coming from, you know, there's shipping involved.

During COVID we faced this a lot where shipping industry just took a big hit and material wasn't showing up in time. I mean, I had issues where generator timelines were sixty to seventy weeks on a project I could build in forty to fifty weeks. And and and so when it's a life science project or a healthcare project, I gotta have a generator, otherwise you're not gonna get

Rishi (37:12)
Yeah.

Tanmaya (TK) (37:19)
permit to use the building. So you can have a building sitting there for 20 weeks because you don't have a permit. Because you don't have a permit, because you don't have a generator. And generator depends on, I mean it it comes from you know, depending on which generator you're buying, it's coming from outside the country. And and so so yeah, it's a hard problem and and I think since COVID contractors, especially the bigger ones,

Have gotten way more sophisticated at it. They have teams that that track critical material at least, like all the key equipment is tracked, they stay on top of it. since COVID, I've done projects where we would say, you know what, first thing we're gonna do is gonna order the generators and air handlers, even though building is still in design. But we're gonna push the design team to lock that in, order that material while you're designing.

So that it shows up on time because I can build faster than this thing is gonna show up. Which didn't used to be the case. Yeah. Like so so I mean, that kinda that's that's that's my take on it. Yeah.

Rishi (38:12)
Mm-hmm.

Dependencies are complex and they can influence your process completely.

Tanmaya (TK) (38:33)
Yeah. Yeah, no, absolutely. And that that's why I started like why construction projects are so complex. It's a spider web of things that can impact it. And and most people remember a project based on that one thing that impacted it. nobody remembers nine hundred and ninety nine things that were solved. but it's it's just i i it all leads to getting your final inspection passed, right?

And it's a laundry list of things like the elevator needs to be working, all the slopes need to be ADA compliant. There are a lot of little things. And even if one of those things is out of whack, you got a problem, right? The fire alarm is like even like I mean again, another digression, but buildings have become a lot more complicated in this day and age as well. Right. Like many people look at pyramids and say, what a great building. Well, it didn't have any H V A C in it, you know.

Rishi (39:27)
Mm-hmm.

Tanmaya (TK) (39:28)
Yeah. As soon as you put mechanical, electrical, and plumbing, that's where complexity comes, right? And there is a lot of software coding that goes into the buildings now. Like a door is controlled by a security system and a fire alarm system because in case of a fire, the door needs to be able to open. But for security reasons, it should only be open when there is card can open when there is card access, right? So there is a software hierarchy that needs to be talking to a door.

whereas back in the day a door was a door. Right, you know, like it's just so so things change with time. and and so buildings also become complex in these little ways, right? Like every room will have its own temperature control or the whole floor will have one temperature control. The coding changes. The the software code behind that changes.

Rishi (40:17)
I really appreciated your door example. The next section, TK, is technology in construction. Over your career, you've seen construction move from manual paperwork to digital tools. Where do you think technology has helped the most and where is the industry still stuck in the past?

Tanmaya (TK) (40:19)
Yeah, yeah.

one of the things I would say, I think construction industry in some ways was an early adopter of CAD. I think it was right up there with airline industry and and others on on but then we never transitioned to using 3D model for permit submissions. Like three D model is something

It's done on large projects for coordination, but it is not the source of truth. The contractually the source of truth is still 2D PDFs. So even though I I feel like industry was probably the first one we started using CAD to build drawings, we never transitioned that to getting out of 2D PDFs being the source of truth. so in the in that case, I feel like it's still stuck in the past a little bit.

the other thing industry has done well is with the iPads and iPhone where you could get a lot more data on your screens. I think industry adopted that well. Like I see a lot of our foreman where I remember back in the day we're like, well how do we

Train a foreman on using a computer properly. And we used to have these things called computer kiosks and BIM kiosks on site where they can open a computer and or there'll be a desktop setup that they can log in and use it. And it required a lot of training. iPad kind of minimized all that because iPad's so intuitive. And more and more tools that that work on iPads and iPhone allow f a foreman who's already used to using an iPad and an iPhone.

really being able to consume the data. So the friction of data consumption I think has been reduced. I can't remember the code, but I think anything that is good has its own bad parts as well. And the the issue is the friction of data flow has gone, but I think that has also led to too much data being bombarded on Foreman level.

Right. We like I I some in some ways, like when I ran projects, I would always say we are not USPS, we are not postal service, right? We are our job is not to move data from point A to point B. Every engineer, every project manager should be filtering the data so that fore mens gets what they need to work. Right? They're they're they're they're getting filtered data. So it's important that just because we have the ability to push a lot of data very quickly to an iPad, we're not doing that. Right? We are.

We are filtering it through as humans. So yeah, that's something I think AI can probably help with more in that filtration.

Rishi (43:16)
Your new company, Melt Plan, focuses on helping contractors send more bids and reduce takeoff time. What did you observe in pre construction that made you believe this was the next big problem to solve?

Tanmaya (TK) (43:31)
Yeah, that's a great question. I mean, I come from ops. and then once you get in an executive role, then you do manage a project swoop to nuts from selling to closing out. And one thing I would say with all my years of experience in op, no matter what you do in ops, you can never outperform the results you can get in pre construction. You just cannot. So again, I use the car industry example. why are they able to

Manage the cost of building a car so well. They spend a lot of time on design, they spend a lot of time on prototyping. They might try 15 different prototypes and then they figure out each and every step and then they push it to production. Right? So most of the time that is spent building a car is in pre-planning. So that is what pre-construction is, right? Simply put, pre-construction is everything you're doing before you put shovel in the ground. I I know traditionally when in construction, you mean pre-construction, people think estimating.

But it's estimating, it's scheduling, it's 3D coordination, it's it's all of that, right? So most successful projects are that have had robust pre-construction planning, because more pre-construction planning you have done, more seamless your project's gonna be. Because when you're running ops, you just wanna let me say it differently. You can screw things up in ops, but you're never gonna perform better than what pre-construction gave you.

That's a fact. And so so things can go bad, but it'll be very hard to do it better than the the bed you made during pre-construction, right? And the other thing I notice with pre-construction, this is why we at MELPLAN we decided to focus on that. Pre-construction is always a rushed process because everyone wants to get to building a building. Once a client has decided they want to build a building.

Rishi (44:54)
Mm.

Tanmaya (TK) (45:19)
They want to build it fast. They want the building yesterday, right? They have done their revenue projection. Sooner they get it, they can get revenue out of it. If it's a revenue building, if it's an office building, still, they want to move in their employees as soon as possible. And pre-construction always seems like a likely candidate to cut time on because it's like, yeah, it's a bunch of people sitting and talking in a room. Like, you know, it's just I'd rather them be building a building, right? So so pre-construction tends to be a rushed process where I feel like.

Okay. There are two things I want to solve in pre-construction. First of all, use AI processes to allow maximum time spent on strategic thinking and planning, then doing you know keystrokes on things that can be automated using AI. That's the biggest one. Second thing is the mindset in pre-construction needs to change because, as construction people, for us, change is bad. Every project that we have problems on is because things changed.

Which is true during construction, but during pre-construction, I think change is good because then you beat up your plan multiple times and you build a more resilient plan that will because you're gonna have things change in construction, right? I think it was I don't know if it was Napoleon or who who said that, no plan survives first contact with the enemy. And and and so so whatever plan you go into construction with, something's gonna change. You know, something's gonna happen and and

So having a more resilient plan is very important. And to do that, we gotta have more change, like you gotta try different things in pre-construction to see your plan holds. so I think now with AI has the ability to actually we have a tool that can read documents, that can reason things. Now you have a co-pilot to these experienced pre-construction people who can help them.

reason through these things, do these changes multiple times so that you make a better bed. So that's what MELT plan is focused on. we are we are here to make sure that we have a AI native pre-construction strategy for companies.

Rishi (47:24)
Yeah, there's a saying the plan itself may not be important, it's the planning process. That tells you so much.

Tanmaya (TK) (47:24)
Yeah, yeah.

Mm, yeah.

Yeah, exactly. Yeah.

Rishi (47:35)
The last section here TK is the future of construction finance. And this is also the last question. Looking ahead, ten years, what do you think construction finance workflows will look like? Will project managers still be coding invoices and chasing paperwork or will technology finally close the gap between operations and finance?

Tanmaya (TK) (47:43)
Mm.

I don't know, Rishi. I think you should be answering that question because I'll tell you my vote. if you can automate the coding process and make a project manager a reviewer instead of the doer. Yeah. Like AI takes first taps, codes everything, and project manager is like just doing like a yes no, maybe, right? Yes, that's right.

move forward. No, no, this is I know this is wrong. Okay, I need to redo it. And then maybe I need to double check it. Right? If if if we can do that and most of the things are yeses, that that would be ideal. In 10 years from now, where the industry is gonna go, actually that if with AI and things changing so fast, I don't know. I think it was like Bill Gates who said that we

overestimate what technology can do in next two years and we underestimate what it can do in 10 years. And so so I I I don't know if I'm overestimating or underestimating. I mean who knows? We might have a bunch of robots building the building in 10 years, right? And and that might change how coding happens, right? so I I I I don't know if that's possible or not. I know

self driving was supposed to be done and dealt with by twenty twenty and we are in twenty twenty six and we have self-driving in like what five cities. So so that might or might not happen. So I don't know about that, but I can tell you today what I want. you tell me w what what companies like yourself can do. Right. So

Rishi (49:37)
So we do a construction accounts payable software and our tech we can automate coding. So the accuracies usually end up being h in high nineties. The project managers take a look at it and they say, Yes, it's right and no it's not right, we need to change it. So we are already there in terms of coding, but getting the accuracy is high, it will require better models, it will require some more time. Ninety is still not hundred, right? The project manager.

Tanmaya (TK) (49:50)
Hmm.

Mm-hmm.

Rishi (50:05)
still has to figure out which ten percent is not right.

Tanmaya (TK) (50:09)
I mean 90 is pretty good. I mean, I I always I mean we at Meltplan, we we call this the 95% threshold for the pre-construction processes we are implementing using our tool. So if any activity we compare against like what gold standard of like this guy has done a great job, if we cross 95%, we feel there is value. 80 to 95 percent zone I call is like the break-even zone.

Less than eighty, I feel like the guy is spending more time f chasing errors and it's actually counterproductive. So at least at least that's that's our philosophy. Yeah.

Rishi (50:42)
You think

So you think ninety is good enough?

Tanmaya (TK) (50:48)
90 is on the upper

end of break even. If you can push it to 95, to be honest, that that's where it's like a big difference. But I still think that if you're at 90, you should probably start putting it in front of because it this is all very nuanced, right? It depends on how you get the invoices too, I'm guessing, right? Is it digital? Is someone writing by hand? You know, like how how are they coming to you? you know, are they scans?

Rishi (51:08)
yeah.

Tanmaya (TK) (51:16)
So so it all it all kinda depends how the data is gathered as well.

Rishi (51:21)
Yeah. Yeah. Actually if a human cannot read the invoice, there's no way AI can read it.

Tanmaya (TK) (51:27)
Absolutely.

That's why I say compare to a human, right? Not compared to an idealistic scenarios where stars align and everything is perfect, right? Compare to a good human competent human. Yeah. Yeah.

Rishi (51:30)
Yeah.

Right. And humans. We we make mistakes. Don't

think like the project manager coding a thousand noise. He he or she is getting all thousands right, you know?

Tanmaya (TK) (51:47)
Yeah.

No, I mean I I was telling this to our CTO that see human make mistakes, but the reason we're okay giving them a pass because they can also explain the mistake and kinda sometimes bullshit their way out of it. AI has at least not learned yet completely. But

Rishi (52:03)
Yeah, yeah.

It was only 10

years ago I was working at commodities trading platform. it was a company, we did commodities like corn and ethanol trading. And half the time our reports were wrong. It says like value at risk is like $100 million, and the CEO looks at it as like, my position is only like for five million dollars. How can we the value at risk be $100 million here? You know, so this definitely.

Tanmaya (TK) (52:14)
Mm-hmm.

Yeah.

Yeah.

Yeah, yeah. All right.

Rishi (52:36)
Reports and a lot of other stuff that humans are producing or the systems are producing, they could be wrong.

Tanmaya (TK) (52:42)
Yeah, no, no, absolutely. That that's why, like getting you know, I mean, d when you're talking to your customers, we try to do that. That hey, send me what you would consider good, and I would like to run AI and compare to the output that you considered good, and then run a comparison with that, right? So so I think I find that to be very helpful. And it's revealing for the company as well to see, okay.

this is how AI is performing compared to what they would consider like top of the line or like not compared to like the top ten percent of their employees.

Rishi (53:20)
Yeah, yeah. T K, I enjoyed this conversation. Thank you for giving us this time.

Tanmaya (TK) (53:26)
Yeah, no, I appreciated it and yeah, have a good rest of your day.