AI is Transforming Drug Discovery. Just Not The Way You Think

Monisha Saldanha (00:00) Today's episode explores how AI is opening opportunities to build software for niche industries. I'm joined by Cole Weiler, CEO and founder of BoltWise, a company is automating quoting in the procurement process. We'll explore how AI makes possible to build software for niche where it just wasn't economic before, how their technology is transforming procurement, and what it means to build and scale in Colorado and beyond. Let's dive in. Cole, delighted to have you here today. Thank you so much for joining us. Cole Weiler (00:33) Thanks, Monisha Really looking forward to the conversation. Monisha Saldanha (00:35) Yeah, me too. What was the inspiration behind building Boltwise? What was the problem that you saw Cole Weiler (00:41) Yeah, so I answer this question all the time. Building really weird technology for nuts and bolts. A lot of the time people ask, how'd you end up doing this? So for me, prior to starting Boltwise I was working at McKinsey where I was a consultant specializing in procurement for industrial companies. And I've bought pretty much anything and everything for a lot of companies you know pretty well. And through that, none of it's easy. Procurement is really hard. expertise in the first place, but notice there's like some categories are a lot harder than others. And one of those is what I'm selling for now. So we work specifically with fastener distributors. So nuts, bolts, washers, pretty much anything holding things together. But it's this broader, it's part of this broader category called industrial products, which could be anything from fasteners to pipes and pipe fittings to electrical connectors to, you know, the gloves that employees at manufacturing plants are wearing. And the reason why these are so hard to go out and buy is they're super old category and it's hugely fragmented category. There's millions and millions of the unique parts, but none of them are really sold on a part number, lot of these parts are being bought and sold on free text descriptions that someone's typing out in an email. And I noticed compared to categories like semiconductors, which I used to buy a lot of all those are stored on really well ordered part numbers and I can go out and I can buy 20,000 unique parts in 24 hours. Fasteners was just never the case. And after doing that a couple of times, we can see with ChatGPD coming out in fall of 2022, realized there might be a new opportunity to go solve this problem that seemed really hard to get your hands around prior. Monisha Saldanha (02:13) And how did you validate that were Cole Weiler (02:15) Yeah, being like probably every other startup founder, just being really annoying and, you know, consistently calling and finding people who are willing to give me time. I was really shocked at how willing folks were to talk to me. I think a lot of it's because they're entrepreneurs themselves. And so it's something that they were eager to chat about. And it's also a problem that's kind of been a facet of the industry for the last 60 years. The one thing I noticed that was a little unique is a lot of time you read advice on, go ask your customers what they want to solve, want you to solve. And I do believe that's like prevailing the correct advice. But when I started asking customers about, Hey, is this something that you guys struggle with? It was more so that this is just a part of being a fastener distributor. You have to quote the components. That's really like what it requires to be able to sell these parts. And so it's kind of a paradigm shift to ask them what if a computer could help you go out and quote these a little bit more quickly. And then just co-creating with those customers. Monisha Saldanha (03:11) And how did you find out that they were willing to pay? Because they were willing to work with you on the solution. But how did you determine they were also willing to pay for the solution? Cole Weiler (03:20) Yeah, start charging people. It was definitely the hardest thing for me too. It was just coming from a different background. And I'm someone who is always looking to find the probably perfect solution. But I think we found to a place where, this is providing a value for people and start charging them and just continue building on that. yeah, I think that's probably the only way to really go about it. that data sets, those data sets to start with. Monisha Saldanha (03:41) Can you talk a little bit more about why it would not have been possible to write software for this industry previously and what's changed? Cole Weiler (03:50) Yeah, it's really interesting topic that's kind of evolved over the last couple of years. And I feel very lucky to have found myself in a place to be able to observe it. If you look at broadly more the technological themes that have kind of played out in software over last 20 years, early 2000s, mid 2000s was this broad concept of like horizontal SaaS, the HubSpots and Salesforce of the world, these large ERPs servicing a lot of different verticals. And then obviously moved into mobile and then cloud, and then it became more popular to this vertical AI, you know, finding these underserved industries where these horizontal products really just don't work very well and being able to say, okay, it requires a lot of domain expertise to understand what these business owners need. I think that has been dramatically accelerated by AI. And if I look at like what we're building specifically, it was an insurmountable problem just a couple of years ago. The way you would have solved this before AI, you could have used something like a machine learning model to be able to solve it. But if you look at how you would solve it, simplistically, you would use something like a rules engine. So effectively a bunch of if statements saying, you know, if I see this letter combination in the sentence, this is what that means. And creating this in another library. The problem is this world is so immense. The amount of nodes you have just continue to grow and the amount of surface area for rules engines is really massive. And at some points it begins to conflict too. So if you told me, Hey, you have a thousand engineers who can go out and write rules engine synonyms until the end of time. Like theoretically you can go on solve this traditionally, but fastener industry, it's about a $30 billion industry in the U S industrial distribution is about $300 billion in total in the U S. And so it's a, it's a sizable industry, but there's healthcare and education and all these like really massive industries out there, but what large language models allow for is really quick pattern recognition. So I don't have to go out and hire 100 engineers to go out and build this really complex rules engine. I'm able to do it really quickly and I'm able to service this problem that if you had tried to attack previously, you would have just fallen on your face really quickly with. Monisha Saldanha (05:50) And are there other niche industries where you've seen software like yours come up? Cole Weiler (05:55) All over the place. I mean, I definitely live in the bubble of vertical AI. And so I see a lot of it of, know, folks I know in my network who working on similar projects, but, and our first investors specializes in vertical AI and vertical SaaS. And so there's folks working on stuff for car washes and veterinarians and, concrete mixing companies and super niche medical tech opportunities and it really has just opened the door for all these companies who got this kind of broad-based solution to get something that's custom built for them. I think the other thing that's really interesting here too is it's allowed some of these individuals, distributor owners, I was at a meetup for fastener distributors last week where we were talking about AI and there's folks in the audience who are building tools internally for their team to go use and that would have been really hard if you're one of those our customer and a smaller shop, you can't go out and hire an engineer to go build software for you for the next year. But now he's the owner of the business and he's able to go out and do it and build a solution for his team really quickly. So I think not only software providers being able to do this, but actual business owners being able to go out and build solutions to these problems that have just kind of been a nagging expectation of the industry have really changed. And that's what I saw, like you said, at the beginning, when I reached out and started talking to business owners it wasn't, we struggle with quoting, right? Quoting sucks, but that's what it is. That's what it's like to do business in this space. Monisha Saldanha (07:20) Yeah, they were resigned to it and it sounds like they're no longer resigned. Cole Weiler (07:23) Yes, and on a lot of different topics, and I'm sure there's so many areas out there. I think it is really cool. You see the resurgence of people who know spaces really well being able to go out and solve their own problems because the barrier to build technology has gone down. Monisha Saldanha (07:38) And is that a threat to what you're building here? Cole Weiler (07:41) Yeah, it's a question I think about all the time and I'd be concerned if I wasn't, but I believe there's a lot of opportunities to go out and build software that that solves problems for your company. But at the end of the day too, there's still a lot of really hard problems out there. This specifically is an incredibly nuanced and difficult problem. Like I said, for better or for worse, I wish you could just take an LLM you know, ship it out there and have it produce great results. But we just have found that it is in the case. And really the symptom that I've seen is the edge cases. So to go out and build a solution that works 50 % of the time is, I do think doable for a lot of people. But to go build something that's production grade. And when you think about B2B applications specifically, the margin of error is incredibly low because, in my space, if I ship a pallet of bolts, which is effectively just a block of steel to the wrong part of the country, is a really bad mistake and it's an expensive return to take back. Not even thinking about the customer consequence of they have a production line that now is down because they don't have the right part because it's a quarter of an inch too wide. So the expectation of accuracy, I think specifically in the B2B space is really, really high. And to be able to get that out of AI and to get that out of software in general, still requires people to spend an immense amount of time dedicating to buildings specifically for that. And the other thing I think about too is at the end of the day, their core competency is being a really good industrial distributor. They shouldn't be thinking about how they maintain production rates offering, know, focus on what you're good at. Obviously fix gaps where you can. but it's not something we're worried about in the near the Monisha Saldanha (09:14) Did you ever face pressure to go broader instead of staying focused on a specific vertical? Cole Weiler (09:19) Yeah, for sure. They're, you know, especially when you look at VC-backed companies, we work with some really great investors and so they've always been bought into kind of the vision I've had for this business. And they've been willing to support me through that. But I think when you look broader at investors, they're looking for the, you know, grand slam. And that's really obvious to do in these really large spaces. So when you go into the vertical AI space, I think that becomes a little bit more opaque and you have to have a really clear vision for why you can go build a company there. My belief is that you have to understand the domain incredibly well and if you fragment your focus across a bunch of different verticals you end up building a relatively similar product to these horizontal SaaS products that have kind of failed customers in the past and so that has always been our objective is focus on the customers that we serve and do a really good job for them and then when the time is right figure out how we can expand. and bring this offer to more customers. Monisha Saldanha (10:16) So did you face pressure not from the people that are invested in you now, but other potential investors? Who was the source of the pressure to go broader? Cole Weiler (10:26) Yeah, I'm sure there's plenty of investors who have passed on us because they're looking for someone who's going to be able to service quoting across every industry more broadly. But I think there's also lot of investors who have become really interested in this vertical concept and being able to master it. I think the other thing that's important there is being able to know what the company's going to look like long term. And if you're not able to solve for that, it becomes really difficult to figure out. Hey, are we going to be able to continue to serve our customers as best as possible going forward? Monisha Saldanha (10:54) And what's the decision that you made early on that felt risky at the time but proved critical to making the business model work? Cole Weiler (11:01) Yeah, I think the riskiest thing was building a point solution. When we started this, was like sacrosanct to say, I'm going to go build a point solution for a vertical that's as narrow as something like fasteners. And the reason being is like a lot of what vertical technology has been built, you know, the previous five years was system records. So building the system record where all the data lives and is the central point. I was against that. The reason being is I believe it's really hard to do that. I think it's hard to replace a system record. There's a lot of nuance that goes into building system records for old industries and it's hard to rip out a lot of our customers. For our space, system records and ERP and their whole entire business operates out of this one piece of software. Asking someone to rip that out and replace it with something else is like really hard decision to make. And when I spoke to business owners, they're like, I'm still growing back hair from the last time I ripped out an ERP. There's no chance I'm going to go work with a startup doing that. And so we took a bat on what if you built a point solution that integrates into an ERP but focuses on functionality that an ERP would have a really hard time recreating. And the reason I believe it's hard time is because the problem is so nuanced and complex that you can't spend partial focus on it. And now I think that's actually a really popular model. The way I view it is actually that AI companies should provide the data inputs and outputs to a system record. And that's where we pitch to our customers. We're not here to replace your ERP. We partner with ERPs and they're actually really great business partners to us. But our goal is how can we get data in and out of your ERP as quickly as possible and use this technology that's available so that you don't have to rip it out. I liken it to like if you're driving an old car and you want to get a touch screen for your radio and you want to upgrade from a cassette player, but I tell you you have to rip out your whole engine, it's like a really hard decision you have to make. But if I can tell you, just plug this new touch screen in and you're up and running in a day, it's like much easier decision. And so we've actually found that to be really effective. And I think it's allowed us to go work with a larger swath of customers that might not have been willing to switch their system record previously. Monisha Saldanha (13:08) And how many different ERP solutions are you integrated into? Cole Weiler (13:12) Yeah, so currently we integrate with three. So there's three pretty big players, specifically the fastener space and industrial space more broadly. And so that's where we focus. We focus on where our customers are. And as we grow and expand our business, we are looking at like, all right, what are the next large players that we could go integrate with? But that's really been our focus. And I think that's a huge, a huge topic that business owners of, you know, not just distributors, but anyone who's who has a system record currently should be looking at too is, am I using a system that's able to integrate with some of these new solutions? Because if you're using an on-prem position and there's no API, there's very little I can do to be able to flow data in and out of your ERP. Monisha Saldanha (13:49) Yeah. And Do you have plans to expand to more ERPs or are you happy with the three that you're integrated with at the moment? Cole Weiler (13:57) It's like you were sitting in our office this morning. something we're talking about all the time. It's really challenging. We have a good team. My co-founder, his background was in building integrations for Uber. so, way Old system records like at Starbucks and McDonald's and stuff that's not even, the documentation's not there and it's really hard to come through. so. Monisha Saldanha (14:06) Mm. Cole Weiler (14:16) He's very good at it. We have a pretty talented team at Building Integrations, but I always focus on like where are the customers, where are people pulling us and supporting those ERPs because it's massive landscape. There's a lot of providers. out Monisha Saldanha (14:29) Yeah, really interesting to think about and challenging. And at a practical level, how has AI changed the cost structure of building software for niche industries compared to even five years ago? Cole Weiler (14:39) Compared to even six months ago. Yeah, it is totally transformed it. And we're as a team constantly thinking about how can we be pushing ourselves? So both on the engineering side, that one's a really obvious, super well documented Claude code agentic coding, giving you crazy amounts of leverage. That I think is obviously allowed you to release features much more quickly. There's unintended risks to that. Something that we're always constantly evaluating. How do we make sure we're still releasing high quality features to our customers? not just a lot of quick features. But I think that is something that's really powerful, being able to out and build some of these niche quests that might've been insurmountable, leveraging the agents in the LLMs themselves and being able to go out and provide workflow support. I think building agentic layers into B2B software is still really challenging. It's a very nascent topic and we're always looking at how we can provide more agentic workflows in our system, but that's obviously allowing you to go capture these really complex decision criteria that, going back to the rules of your approach would have been really fragile and brutal if you tried to do it traditionally. And then the last piece is from a non-engineering side too, we see a huge amount of leverage. know, Our sales team is able to do a lot more sophisticated outreach than we would have been able to do before. I'm able to do a lot of stuff that I wouldn't have been able to do before just because having to prioritize where I can spend my time. Now I can have an agent go run out and fill out a task that I might have previously just said, it's not worth it for me. Monisha Saldanha (16:02) Yeah, and you're using AI on both the delivery of your product but also the production of your product. Cole Weiler (16:09) Yep, yeah, we're big believers in both of them. So yeah, we run a lot of our product through providers, mainly OpenAI and Gemini are the two providers we use most in our production environment. And then our team uses a mix of all three major providers for day-to-day tasks, building out code, like I said, some of these non-engineering workflows. But yeah, we use it on both sides pretty heavily. Monisha Saldanha (16:33) Great. And can you walk us through a real example how BoltWise uses AI to solve a problem that would have been too expensive to tackle before? Cole Weiler (16:41) Yeah, so one, and this is something too that I think is a really interesting concept, but one is OCR technology is mind blowing. I, even since I started this business in 2023, it's a leaps and bounds better than what it was. There are companies who have spent their entire existence working on building out really good OCR capabilities. That is a core aspect of what we're able to do. Now, obviously there's nuance to being able to deploy it correctly for the industry we serve, but we have a five person engineering team currently. That would have been a massive lift and I would have really sacrificed future richness elsewhere to go out and make sure our customers were having a great experience when we're reading in these PDFs. Now we're able to go out and use some of this technology off the shelf and like I said, provide our own data and provide our fine tune it so that it works well for our application. But even when I started this in 2023, it required a lot more engineering effort than it does today. Now it's like... it's definitely still, like I said, not trivial, but it's become significantly easier and I can provide an experience that's on par to these companies that have spent their whole existence building just this one piece. So that's, think, a really obvious one in our product. And then on the engineering side, I think where we've seen the most is on the coding. I just said that the non-engineering side is pretty incredible too. Before this call we were talking about being able to update HubSpot right after we get off a customer call and having pretty much 100 % fidelity to your notes. But on the engineering side, to see how quickly we can release features now that would have been weeks, if not months, to work on is absolutely a game changer. And our customers recognize it too of how are you guys coming out with features so quickly? One we have a really good team, but two, it's because these tools just allow you to move so quickly. Monisha Saldanha (18:23) Great. What advantages or constraints have you experienced building BoltWise in Colorado, especially as an AI first company? Cole Weiler (18:32) Yeah, it's a good question. So I think I'll start with the advantages. One is I think we were able to build a vertical AI company because we have really great investors and one of those was Range who is a prominent seed stage investor in Denver area. And what I think is a benefit is We're building for a very niche industry, but I was able to you know, to have conversations with Chris and Adam at range and show them this is what I think the opportunity looks like. And this is why this is something I think could be a durable business. And I don't know if I would have that same opportunity if you're on the coast where there's a, you know, a plethora of AISDR companies or AI coding agents that exist. It's kind of harder to make a splash with something really niche like fastener distribution, for instance. The other aspect is Colorado for any founders in Colorado, there's a great program through OEDIT, which is Office of Economic Development for Colorado. And so we were awarded $200,000 through their grant process there, and that's focused on businesses, not just technology businesses, but a lot of physical product businesses too, who are building new technology in Colorado. And so that's a great way to get non-dilutive capital and be able to grow jobs. think we've like, We went from me being the sole employee in Colorado to now we have eight employees in Colorado in just a short amount of time. And a lot of that was just, you I felt invested that I should be building this business here. And I love it. We love working in person and being able to collaborate here. From a disadvantaged side, I think, you know, it cuts both ways. At the same time, you're able to go out and work with really great investors in Colorado not just Range but we have a lot of investors that are located in the Rocky Mountain region. At the same time, there's some discrepancy with being in the coast, being that close in proximity to, whether it's SF or New York, being that close in proximity to the capital. The hardest thing I think I've realized though is the density of tech talent in Colorado is obviously pales in comparison to what you would see in New York or San Francisco. We do have really talented engineers here, but we spend a lot of time looking on how we grow our team, and most of our team actually isn't based in Colorado, which is for the engineering side, which is something I wish we could change. Monisha Saldanha (20:45) And What advice would you give to founders in Colorado who want to build AI-driven companies in less obvious or unsexy niche industries? Cole Weiler (20:53) The one thing would be definitely if you're building here, checking out OEDIT is a great place to start. mean, and just to give them a shout out, I think it's awesome that our government is supporting people building businesses in the state. I think it's a really great incentive and I am really grateful for that. The one thing I would say too though is it is important to be close to your customers. So most of our customer base is in the Midwest, industrial manufacturing, super heavy, actually pretty slim in the Colorado area. If I was to go back in time, like that's really where I would have made a change is I love Chicago, I love Denver more. I think there's great talent, but like most of our customers are in the Chicago area. So being able to be close to those customers and collaborate with them, because a lot of it's just really getting out information from their knowledge of this industry and being able to encode it into tech. But our first customer, my co-founder is in San Jose and our first customer is in Sunnyvale, which is right down the street from San Jose. And so we spent a lot of time at that customer's office, building with them and asking their team questions. And so obviously, proximity to your customer is going to be most important, think, especially in the early days. Monisha Saldanha (22:01) What's an example of a customer industry where BoltWise has made a tangible difference that wouldn't have been possible pre-AI? Cole Weiler (22:08) Yeah, I mean, I think with most of our customers, were, we're allowing them to do things outside of just responding to quotes. And so the most common example is, you know, with these ERPs, it's forced people to do a lot of data entry. So if you look back into like early 1990s, before any of these ERPs started popping out, people were writing quotes down on pieces of paper and maybe they're faxing them or maybe they're just storing them in a file cabinet or a card system. But with ERP's, a lot of it's moved to typing in on keyboard. And so there's a really special moment for me and my team. We went out and visited a customer and we were sitting there with one of their sales managers and he said to me, you don't realize how many people sit in the parking lot every morning, dreading coming into work. And I really hadn't realized that. I thought like, okay, this is a great way for businesses to be able to service more customers. That was all my focus was like, how can you respond to quotes more quickly and more business? But for a lot of people, they got into sales because they liked to interact with customers. And the job quickly evolved into spend as little time on the phone as possible and then spend the rest of the time typing in the order. And I think that's really where we've noticed is freeing up that time to go actually do the sales part of the job. Getting away from being an order entry specialist and more into actual sales reps. And a lot of times that's what people like to do. They like to solve problems for their customers. They like to interact with their customers. They like to be a resource that has expertise. And when you're typing in POs all day and you're typing in quotes all day, it gets away from that. So, and maybe I'm being like a little naive or optimistic on this, but I see that to be, and I think about it a lot of like, what is the impact that AI is going to play out on society. And I think there's obviously a lot of risks that come with that. But the one thing I've seen just from work we're doing is you give an opportunity for people to get out of the data entry mode and get back into the thing that makes them uniquely human, which is usually some type of interpersonal interaction or relationship. And so I do get a lot of hope that if you can free up people to go do those tasks, like a great example that we have customer and his objective was, I just want my sales rep's calling. I just want them calling and asking customers what's going on, how their day is, what project they're working on, what their next six months looks like, are they excited about a new project coming up? And his team told him, we just don't have time to do that today. And so having been able to give them time back to do that, I think allows some of these smaller businesses specifically, most of our customers, SMBs, somewhere between 10 and $100 million revenue, they can now go to... to compete and free up their team's time to go do these things that they're like resource limited to do previously. Monisha Saldanha (24:35) that's so exciting. What's the mistake you made early in building BulletWise that taught you an important lesson? And how did it change the way you lead or make decisions today? Cole Weiler (24:44) How much time do we have? That's a hard question. I mean, I think anyone who starts a business would probably be, I'm sure you probably, as you with this, you've seen the mistakes are really easy to come by. I think the one thing that I regret is not being aggressive enough. I think I alluded to it early on, but. I really want to make sure what we're doing for our customers is going to service them well and they're going to see value out of that. And I think I underestimated the value we were returning to our customers for a really long time. And what that limited was not only how hard we pushed the product, but how hard we were able to go and scale it to other customers. Someone give me some advice and what they said is like, you're never going to be satisfied with where your product is. I've slowly realized that's going to be true. But that is something I wish I would have, you know, focus a little bit more hard on was how can we go and scale this in a way that allows the customers who are getting value to get access to it quickly, rather than, you know, constantly waiting for, okay, as long as we make this one new feature, now we'll be ready. We've been able to shed that, but it was a really hard lesson to unlearn. But yeah, mean, plenty of mistakes in between, even I've, you know, making mistakes every day. And I think that's the part I actually really like about this is I love learning. And this has been a really great opportunity to just learn at a really fast pace because you're just making mistakes so quickly. And the, and the feedback loop is actually pretty fast. You have a really good idea when customers aren't buying your software, when people aren't returning your calls or, when you're not getting things out on time that like something's broken there. And it gives you a good opportunity to go dig deeper and figure out, okay, what were we doing wrong and how do we go fix it? Monisha Saldanha (26:22) Great. And last question for you. Cole Weiler (26:25) I'm curious what your question is. I see your bookcase and it looks pretty impressive. So I feel like you have a good answer there, but yeah, I actually, most of my favorite books are not, I like a lot of nonfiction, but most of them are not business related. I like just to learn a lot about the world in general. But one that I really liked, I think for that reason is a book called Principles by Ray Dalio, the founder of Bridgewater. And it is a dense book. He's an intense guy. But I think what is really interesting to me is how he frames these principles on how he runs his business. I think that's really important as a business owner is to know what's important to you in the world and how you want to act and be seen by your team and by your customers. So some of the concepts like having an idea of meritocracy where people's voices are heard because it's the right answer, not because the most senior person on the team said it. I have no problem telling someone that I'm wrong. I would actually love for people to tell me I'm wrong more often because I really just care about getting the right answer for our customers. And the other piece that think really stood out for me from that book is he's big on mental models and I'm really big on mental models too. And that's why I think reading in general is great is because it allows you to build really robust mental models. And I think when I look at what it takes to run a startup, it's a lot of really quick decision making. And it's very taxing if you don't have concrete business, I mean, mental models to run on. And so I'm always very curious about, you know, what are the things that I can learn that I can go deploy in different parts of my business without having to put an immense amount of problem solving and starting from scratch each time. So it's a really good book. He has really good writings in general too. I think he's an incredibly intelligent individual and really perceptive. Like you said, it's a little bit dense and it's got to be something that's up your alley, but that one I really took a lot away from. Monisha Saldanha (28:05) Thank you. That's one I haven't read, so I'll put it on my list. Add it to the bookcase. At the moment, I'm reading the Culture Map, which is about the differences in doing business in different parts of the world and how different cultures have different ways of expressing themselves and interacting. Cole Weiler (28:09) What would your answer be? That is really cool. I was just talking about that with the customer. A lot of our customers just interact globally. I mean, that's like something that's really changed. It's being able to sell globally and the way buyers are buying, the way companies operate in different countries. It's so different than what U.S. business people are interested in or how they operate that I think it's probably pretty well missed. I'm not as experienced with it, but that sounds like a really insightful read. Monisha Saldanha (28:46) Yeah, it's great. Definitely add it to your list. Cole Weiler (28:50) I will. Yeah, that's last thing I need is new books, but I will definitely do it. Monisha Saldanha (28:55) Well, great. Well, this has been a wonderful conversation, Cole. Thank you so much for joining me today. Cole Weiler (29:00) Thank you for giving me the time, Monisha. It's really a pleasure. Monisha Saldanha (29:03) And I'd like to thank our listeners listening to this episode of Colorado Tech People. If you this conversation on how AI is unlocking opportunities in niche industries, consider sharing it with a founder or builder exploring similar ideas. Don't forget to subscribe so you can hear more conversations with innovators shaping the future of technology in Colorado and beyond. Until next time, keep building where others aren't looking. That's where the biggest opportunities lay.