Griff Norville, Head of Technology Solutions – Hamilton Lane (EP.72)

[MUSIC] I'm Scott McDonald and this is Investment Management Operations. This show explores the inner workings of some of the most innovative institutions in the industry. Through in-depth conversations with leaders across operations, compliance, legal, finance, and technology, you'll hear firsthand how key operating partners run their business in an ever-changing and complex investment landscape. You can join our mailing list in access capital-allocated content at capitalallocators.com. [MUSIC] My guest on today's show is Griff Norville. Griff is the head of technology solutions at Hamilton Lane, one of the world's largest alternative asset managers. Today's investment leaders are navigating a technology landscape that is moving faster than ever. From AI adoption to tokenization to digital investment workflows. I sat down with Griff to explore how Hamilton Lane has built a technology and data infrastructure that is reshaping how allocators think about private markets. We cover how Hamilton Lane developed its cobalt platform. Why private market data is still a frontier problem? How AI is beginning to change the way allocators analyze and act on information? And what leaders that asset managers and allocators should be paying attention to right now? Please enjoy my conversation with Griff Norville. [MUSIC] Griff, I'm excited to have you on today and talk about the state of tech and private markets. First, I'd love to go back and hear how you found your way to Hamilton Lane. Hey, Scott. Great to be here. A lot going on in technology these days. A lot of ground for us to cover and things for us to talk about. Hamilton Lane is an asset manager that's focused on the private markets. We manage a big, funny number. It's hard to think about we crossed the $1 trillion mark, which has been really exciting. We've been in business for over 30 years. We've got offices all around the world. I've been here for the last 16 years. So it's been a really fun journey for me. I joined when we were about 120 people or so. And now we're closing in on 800. And I've seen a lot of growth in the private markets. I've seen a lot of growth within Hamilton Lane. Backing up a little bit. And the story of how I got here. I was an engineer as an undergrad. Was put in that place because I was advised by my family, my dad. So mechanical engineer. I was raised around engineers. I was taught that engineers were problem solvers. Go be an engineer. I really wanted to go into business and be an investor. Be a deal guy. So the story of the early part of my career was continuing to try to position myself as a deal person. I came out of engineering school, went on to be a management consultant. Did that for three years? Lived on the road, worked with Fortune 500 companies, solved a lot of problems that they had. Went back to business school, studied alternative investing. This happened to be during the credit crisis, where it was a really good time to be a student. And not directly exposed to all the insanity going on. Coming out wasn't the easiest thing. Knocked on a ton of doors. It was given a great opportunity to join Hamtalan and Philadelphia on the investment team. They learned I was good with numbers and data. I kept getting projects, which is the story of the early part of my career, to work on building the models, to work on building the benchmarks, to be the quant in-house. And it wasn't until I learned to really lean into that, that my career started going in a positive direction and doing things that were interesting for me and unique for the firm, creating value around here that was differentiated. Over time, moved more into the textile. That's the story on how we got here. Does it still feel like an advantage to have that deal mindset intact? Totally. I do get to flex it. We have a ton on staff here at Hamlet's Alain these days. The best technologists we have are super curious about the business. They want to learn about the investment process, what makes a good investment versus a bad investment. You can see in the public markets how the quant side and the technology side is taken over pod shops and quant trading firms, the technologists and the quants want. And a lot of aspects of that business here at Hamlet's Alain we've always supported those types of individuals joining our team. And for people to come in with that unique skill set and then be very curious about the business of investing, that is a terrific combination. I've also seen it happen the other way, where classic business, mind-date and investment types are not afraid to get their hands dirty with tech. That can also be a super successful combination. At what point did Hamilton Lane as an organization go? We have all this information. We should really find better ways to use the information that we've acquired. When I joined in 2010, the idea of having a lot of private market data was a strange concept. No one had any information. No one had any data. There was this old system venture expert that was available through Thomson Reuters that had a bench park you could use. Part of your job in using that data set was to understand its limitations. When you saw weird data, you knew not to rely on it so much. What we ended up realizing is that we were running advisory accounts and separate accounts for some of the largest investors around the world. We said internally within the four walls of Hamlet's Alain, we have more data than anyone else. We should learn how to leverage this. Let's produce our own bench work. Let's build a proprietary forecasting model based on our own data. As we started doing that, it was a virtuous cycle because clients started coming to us and working with us because we shared a like-minded focus on using data to make decisions. It's a cliche now in the marketplace. We were one of the earliest ones to really dedicate ourselves to using private market data to inform how to build portfolios and think with that same mindset that is super popular. It has been for a long time in the public markets about how to use quantitative analytics to think about building a portfolio of benchmarking correlations, how things put together. We found a lot of clients that valued the way that we thought about the world as you attract more clients. You get access to more data. Those things started building on each other and we could dedicate ourselves to hiring data scientists and data engineers to continue to piece all that together to drive my advantage. The topics that you talk about on innovation is that it's own entity or you cross pollinating with other parts of the investment team. There's a lot of different folks at this firm that focus on technology, data, innovation, where I play most of the day. I've got a few different aspects of my role, which keeps it fun. One aspect is I run a business, we call it technology solutions. Technology solutions offers a product and a service around that product. That product is called cobalt. Cobalt today gives our clients access to Hamlet Lane's database. That could be track records of managers and the research that we've done on these managers underwriting of their funds. It's also the 30,000-foot view of the trends in the market. Vinchmarks turning the private markets into indices that look like public market indices, thinking about liquidity and risk. It's also a place for you to connect your portfolio and think about how you benchmark within your portfolio compared to the private markets compared to the public markets and how you forecast and solve for commitment plan. It's pre-investment, it's post-investment, it's high-level market analysis. We also provide managed services around that. We can help you collect your data, make sense of what's going on in your portfolio, because that's still a huge problem in the SaaS class that we're trying to solve in a number of different ways. I would call that my day job. Continuing to grow that business unit, there's 85 people at Hamlet's Lane that work on it now. It's been super fun drawing that from a team of two to 85. I also get to do a lot of fun things in support of innovation at Hamlet's and Lane. One would be working on what we call HL Innovations, which is taking our balance sheet capital, putting our money where our mouth is, investing in early-stage startups. This is not our client capital, which absolutely has access to venture and growth investments. This is really strategic use of our balance sheet capital. To support companies, you'll see a lot of these companies listed on our website. An example would be canoe, Daphne 73 strings, finding these companies that are changing the way that this business works in a positive way, making investments in them in a early stage, and then using our voice, sitting at the table, sitting in on board meetings, whatever it may be, supporting these companies as they grow and make an impact. That's a huge part of what I do. We would be a mess not to mention AI, tokenization, and digital assets. I'm certainly involved in strategy and implementation of those things around the business. I work very closely with our classic enterprise technology IT department to make all that happen. I also work with business leaders around the firm that, as I mentioned, a skill set that you need to unlock at Hamlet's and Lane is to have one foot in business and one foot in tech. We require all of our managing directors to be fluent on tech and to be pushing their team to drive and advantage through tech. It's something everyone here in part must focus on. We're interesting to kind of go back. The cobalt story, could you tell me how that all came to be? I would. part of building our first benchmark around our data. And at the time, we're printing that out in Excel and PDF and emailing it around. And we're trying to figure out how do we get our clients more direct access to our information. They were hungry for it. Their of course was a debate in the hallway. As this cannibalistic, giving our clients all of our information, what are they going to hire us for? Fast forward real quick, giving access to data, to text or clients has been super valuable. It's deep into the relationship that we have with clients. And I think it's attracted a lot of new folks to work with Hamilton Lane that might not have before. So that turned out to be positive. But at the time, there's a lot of uncertainty around that. We thought we're going to take the boldly, we're going to look to maybe disrupt ourselves, give our clients access directly to data. There's one problem. We don't have any technologists on staff that know how to do this. So we looked around about who can we partner with. At the time, I went out to super return in, I think it was San Francisco. There was this cool company there that was building an analytics product around the private markets. They had scrapes and data from publicly available sources. They had a fish ball on the table, winter free trial. I put my card in there. I get a call the next week. I want a free trial to this system. I'm sure everyone that put their card in that fish bowl probably got a similar call. I felt special though. As soon as I logged on and started looking at the analytics, I thought this is it. There's just one problem. They're using all this public data. What if we put our own data in here? And I called up the sales guy that had emailed me the free trial. I said, let me talk to the CEO. We were really up front and aggressive from the first minute. We want to make an investment here. Let's build something together. Ultimately, and I'm compressing a lot of years here, but we built and launched cobalt together. It started as an investment in outside company. And it evolved to us buying our partner out and bringing the core technology in house supporting it from the inside, putting all of our data through it, not being afraid to disrupt ourselves deciding on other boldly. We're going to hire 85 folks around here that don't look like anyone else in Hamilton Lane. These are software developers. These are salespeople that know how to demo and sell tech. These are support folks that turned out to be a good decision for us. I will tell you the first couple of years, the success was not overnight. It was tough to get this thing off the ground, but that was how we arrived at having cobalt. You would think that as a universe of people that are looking at the data, getting them to understand what am I going to do with this. Now I have more work to do. It's almost like an adoption of a new skill set. I've seen a lot of change in this industry. We're going through it right now with AI. We're having to adopt new skill sets. Think about using technology to make decisions, using data to make decisions a decade ago, two decades ago. Not a lot of decisions were made using data and this industry. It was a handshake industry was built on relationships. That's still a super important part of how fundraising, having private markets are sold, not bought, is the common cliche. We've got a bunch of people on the road telling our story, meeting clients and helping solve their problems in a very personal way. All that can be true and it can also be true that clients are jumping on our tech and do it yourself type way, running the same models that used to be perceived as internal proprietary. Only we can operate and run them. Now they're on there running our forecasting model. Our job is to keep that up to date as new strategies come online, become popular as we think about what's happening in the macro environment. We're making adjustments and now it's a conversation with our client using the tech. They developed new skills and had to use this tech once they had direct access that we have internally with our data. We learned a new skill about how to talk to our clients in conjunction with tech, the home market has been evolving that direction. I do want to talk a little bit about AI adoption today, where you guys are and get a sense of where the industry is going. 2024 was experimenting. We were maybe looking at some demos of things. All these demos were shiny, but then you put your hands on it. Didn't they work that well? We were experimenting with co-pilot, which was just okay. 2025 was not procurement for us. Things started to change pretty rapidly, especially later in 2025 in the capability set. We jumped in with two feet, pretty early, late 2024, early 2025 with a number of tools. There are a wrap around the core models. What some of these applications do is provide workflow tooling, allowing you in a low-code way or even using natural language. I want to write a little application to every week, summarizes every meeting that we've had around the world talking about technology. I can write that using that sentence, create an app that I can run every night, every week, whatever I want to do that reaches into our CRM extracts all the information goes into Slack, goes into my email, aggregates it all, and organizes for me every conversation you've had around tech. It's pretty incredible. The possibilities there are endless. We only have access to these tools because we jumped in over a year ago and we said, "We're going to onboard these things. We're going to have a few of them online at any given time. We don't know the one that's going to be at, we're going to connect in all of the places where our data is. SharePoint, Slack, Outlook. We're going to see what happens towards the end of last year. We started writing internal hackathons. It was a centralized decision to get access to these tools. It became a decentralized method of developing applications using the tools. Any business person at Hamilton Lane is capable of building an agent or an application using this tooling with the underlying access to the data we have. It just started to take off some pockets of the organization you would have an all-star here and all-star there. We would put those people on a pedestal, we'd give them a nice little bonus. This is the activity that we want everyone to emulate. All of a sudden, we've got hundreds of agents that comes with its own challenges. That decentralized method allowed us to level up very quickly. Now there's multiple teams that can tell you that they had a four-hour process, taking an initial screening of a co-investment where you would take down some documents from a portal and have some reference calls, do some Google research, four hours of work, six hours of work, condensed to 10 minutes. Now you've got an HTML link you can send out to an MD. It's a four-pager, largely put together with your guidance to the application on how to build this thing and how we think about the world. Also with the vote that we have, around the data we have and all the unstructured documents to turn into great output. You combine all those things together and you turn four to six hours of work to 15 minutes. Then we can spend time on a higher level function, not just the data translation. And on a co-investment where your timing is of the essence, that's gold. We are now moving well beyond getting more efficient. Now we're thinking about things that we could do that we just frankly could not do before. Every time a GP report hits our internal shared documents library, we want to summarize what's going on at every single company within that GP. We want that to plug right into our data fabric and be referenceable by any team that is looking at a secondary transaction or a direct transaction. They would have proprietary knowledge and information on what's a driver within that company, the operating metric level and what's the trend been over the last three years. That's something that we could do with a lot of legwork, opening up a lot of documents, spending a lot of late nights. And it's something that we might do. Certainly if we found a potential secondary portfolio opportunity to be attractive enough to spend a lot of time on it, that kind of thing can happen really quickly. We can start pushing some of these insights out to our clients. But if you want to have a rapid view on the value of your funds or the value of the companies within your funds at any given time, what if you want to more consistently evaluate your options on the secondary market? Each of those things was theoretically possible with a lot of work, but now is more active on going decision that you can have a dialogue that you can have with your portfolio, with your data. You might have a little bit more informed decision making where under the old ways you would look at something, it's really not in scope. So we're going to pass on it without even looking at it. Where now, under you could actually look at it, get a little bit more informed, not burning a ton of time, massive savings there. So we're going to look at more deals. We're going to do more. We're going to develop stronger insights across more sectors. We're going to be able to just level up our game across the board. Come back to your investments in things like canoe, daffney, 73 strings. What was exciting about those? We live in a world that's still quite backwards in the private markets. We're all exchanging PDFs. We tell each other to go log on these portals. It's an awful experience. Go log on a portal, pull down the PDF, read this thing. AI provides a fantastic opportunity to make sense of this stuff. canoe is solving a problem that we've got thousands and thousands of client positions that we have to track. Cashlenose is coming every day across our client portfolio. It was quarterly capital partner statements, quarterly financials coming from the GPs that we have to translate and turn that unstructured information and destructured information for our database, which feeds into our reporting. Canoe can handle all that. They can automatically reach into the portal. They can pull down the docs. They can rename the documents. These documents are named all different kinds of things. You don't even know what they are. Canoe can figure out what it is, rename it, organize it for future reference, extract the data, and then show you exactly off the doc where that data point came from. Click on that data point. Where did that come from? Here's the highlighted section of the doc. Fantastic. That is solving a problem for the world we live in today, where ton of PES flowing and that world is not going away anytime soon. It's becoming more and more of a problem, and Canoe is allowing us to dive super deep into the documents in a way that we couldn't before. Now that more of that extraction is automated, we can dive in and extract all the necessary detail we want about any company listed within that financial. Any mention of carry and management fee, we can segment all that, Richard Data to Study. Daphne is trying to do something a little bit different, and if you focus canoe on the post-investment world, think about all the data transfer on the pre-investment world. Some with Sunderler, GPs are out there fundraising. They're sending out information related to their fund, their prior track record. A lot of that's coming in a PPM or a pitch deck. The GPs are facing a new problem where the wealth market is growing and is super important to their fundraising. And there's all these new marketplaces to work with, and RIAs to work with. It's not as simple as going to your dozen favorite, giant pension plans and getting that check, and calling it a day. The timeline is extended from six months to three years. It's a lot more people you have to work with. So I need to keep everyone informed. Daphne will digitize a GP story for them, all the data that they need to share during their fund racing, and give them the capability, the dashboard, the ability to permission that downstream wherever it needs to go. To Bloomberg, to Hamilton Lane, to any of Hamilton Lane's competitors, to these marketplaces that exist. It also allows groups like Hamilton Lane to go get that data from GPs. GPs will get a request from us. Here's the Daphne link. Take your PPM and pitch deck, drop it into Daphne. Daphne will extract all that data, organize it, ask you if it looks correct. You can edit some things if you want, and you permission that downstream. The beauty is the data is still completely under your control as a GP. If you want to reuse that profile and those data points to send to another LP, go right ahead. So Daphne is building this network of LPs and GPs communicating. And that to me is another future view of how this industry could work. Maybe more optimal view of digital data transfer that doesn't rely so much on the PDF. I still harp on this, that the PDF was invented in 1993. And here we are still reliant on it, trying to build that two-sided marketplace. It sounds like Daphne is trying to crack that code on all that information. The two-sided marketplace is classic cold-star problem that we're removing ourselves through and figuring out. We've got a great consortium around the table. Daphne leading GPs in the market, leading distributors. If we all decide enough is enough, and we want straight through data processing, I do think we can make it happen. That's the Daphne story. You also have a question about LPs. They want more detail than they did before. And this has been an ongoing story in the 16 years I've been in Hamilton Lane, but LPs are asking more and more questions about the companies. And I keep mentioning company-level data. They want to know the operating dimensions. What's happening at these companies? The GPs, of course, want to know that too. They're monitoring their investments. They're trying to make decisions about selling a company, what to do with that company long-term. The GPs have a problem about how to collect that data and then how to value that company. They're going through a process themselves. Go get that data from their company, get that in, come up with valuations, and then in this world in whichever green vehicles are becoming more popular. Valuations are expected on a more frequent basis. We've moved from everything is quarterly to monthly valuations and daily. Many products today are launching with daily valuation. This is where 73 strings come in. A lot of this stuff with canoe with Daphne with 73 strings is about digitizing and structuring that data and developing the workflows for the modern version of private markets. With 73 strings, they help a GP or organize their company level data, help a GP work through the valuation workflow, and then get that in a repeatable motion that is trusted by auditors in order to print that valuation. I don't think we can go to a world where we're all on monthly value or daily value for these evergreen products unless we have solutions like that in place. You figure that there's so much traffic. GPs are trying to work with their port codes. You have that time lag. The LPs are asking questions, but if you can create a single pane of glass, we're good. We're signed off. The LP can actually pull that with little interaction. The LPs are always saying, hey, tell me about what's happening. Is there going to be another capital call coming? And I think that time compression is going to be crucial for growth. Let's condense that timeline. Absolutely. We're all talking about the 401k market. There's a huge potential there. And it makes total sense. Pokes that are investing there 401k. Don't necessarily need the liquidity that the public market provides them. And they could use the enhanced returns that private markets would provide them in exchange for less liquidity. But there's too many things about this market that just will strike people as odd, delayed, reporting, all that time lag that it takes from data to go from company to GP to the valuation over to the LP. All of that's got to shrink massively. Do you think the extension of the private markets will have to be a nevergreen strategy or interval fun where you're going to have to click by or will still be handing maybe a digital version of subdocs? It's a great topic. We have invested in improvements on both angles. We certainly have been supportive of the digital subdoc process. We've been very supportive of improving the process of KYC AMO. For anyone that's filled out these subdoc documents, set aside your Saturday. Sometimes the stuff can be incredibly annoying. We all want to get rid of that process. And we don't think we can truly scale the industry operating the way that we do today. Digital subdocs have been a great step in these marketplaces built around opening up access to the asset class, to the RH channels, and the wirehouses have been a huge help. You've also seen us take more experimental. That's form partnerships with groups that are offering tokenization real world assets in a tokenized form. Digital world world assets. You've seen us partner with groups like Securitize to offer a tokenized version into some of our classic funds, like our Co-investment Fund. And that theoretically will attract a new type of investor. It might be formed a lot of folks that are old schoolers in this asset class. The DeFi community continues to grow when there's certainly a lot of capital out there looking for interesting opportunities. Might attract folks that otherwise would not have come in to this asset class that are more focused on the DeFi world. That's one piece of it. There's just a ton of operational value you get, either from tokenization or even installing blockchain as rails, which is another thing that we're doing here. We can lower the investment minimums to get into our products. Why are the investment minimums so high? To go through that subdoc process, to go through KYC AML to manage the reporting, huge time suck for a lot of people. And there are a lot of groups involved with reconcilations, the fund ad men's, the custodians, and the transfer agents. There's a lot of business built around this. The business collectively is deciding, it's better for everyone, and it's a natural place that it has to go to put this thing on the blockchain, perhaps. And that's going to allow us all to do it better, faster cheaper, therefore lowering minimums. That's just the start. If you think about some of the cutting edge ideas, building model portfolios, making one investment into a model portfolio that then is split between five underlying funds that are private market funds, and then those are rebalanced over time. You start thinking about the number of trades that we have to process and the transfers. The complexity just goes out the window. You can't do that with the old style at all. They will completely limit our ability to attract retail capital. To give them solutions that make sense for them. But in a world in which we've got our investments on the blockchain, you open up this new world of efficiency and you can capitalize on those ideas. It won't really be a game changer for the future phase of growth. Is that a GP constraint or accelerator? I'm envisioning where I have a fund. I don't want to tokenize it, but then I need to bring in my administrator. They need to be able to understand it. I need to bring in my order. That is going to cut down on all the processes and steps that have historically slowed us down where I've got a document. I've got an AMLKYC. I got to clear that. That takes a week. Then if it's tokenized, then if I'm doing secondary, you know who owns it. If you're auditing that piece of it, everybody can look at the same ledger as opposed to back to the deal room. Here's all the documentation. It's all fragmented. It takes time, you can compress time, cost savings, fund expenses. And mentally with that, who's going to drive the change to make that happen? We're seeing GPs drive change here. Many of the larger GPs that look a little like Hamilton Lane see the benefits of being tech forward and leading edge on this stuff to get access to a newer type of investor that wants to access these markets. There's being changed driven by the banks, the wirehouses. Many of them are starting to say in order to work with us. You have to use one of these digital channels. Otherwise, the complexity is too much. We're going to pick you not just to be on our platform to be within our channel, yes, because you're a great brand and you've got a great track record and you've got Trevick Dealflow. That's also important. We're also going to prioritize the operational side. You got to align with this view of the world that we're about into in order to make things easier. They're going to be a big driver of this too. Some of the people that you might think would be less supportive of this tech. We are finding digital forward fund admins and transfer agents, simulacastodians, and groups that are willing to push the envelope, try new things, be the next generation of leader here. If you want to dig in and stick to the old way, you're going to have a much tougher time winning business. I want to talk a little bit about budgeting. There's so much focus on AI adoption, the risk of burning up the budget. I need a bunch of things. I need tools. There's a bunch of cloud storage. How do you think about budgeting of AI adoption and then trying to build integrations and workflow tools for the investment and operations people? First and foremost, yeah, that's a leadership team that's supportive of spending a bit of money here. Our personal experience is that the first step with AI, the savings don't come right away. We had to expand our technology spend budget, jump in with two feet, onward these new tools. It wasn't until some time in that we started seeing the potential for savings in terms of time save. Ultimately, the story around efficiency, it's there. I come back to all the new things that we're going to do. Like any other technology change, it turns into an arms race and new things that are going to add value in order to win the next piece of business and do the next great deal. We're spending more money on tech than we were before. That's part of it. Will we always have three or four different applications that have access to agent building tools? Maybe not. In the early days, it's tough to tell who has the advantage and who's better and that shifts every three months. It's helpful to be spending a little extra. Ultimately, we might need to reconcile that budget down. Will we always use SaaS tools and be potentially at risk for them raising prices ongoing or should we build our own proprietary stack when we have direct access to the LLM? That's another question that's constantly in the hallway. Getting direct access to these SaaS tools or using them rather allowed us to be quick out of the gates and build a lot of great stuff that I'm not seeing other people have access to yet. We might have more control over everything that we deem important around security and these tools are very strict around security. As far as we pick them, the regulators might have other requirements for us moving forward. This is a regulated industry. Those rules have not all come down the pipeline yet. Everything's been moving so fast. We might need to have more direct control. That might mean that we need to go higher, more AIML engineers and security folks build all that up. When you have access to these SaaS solutions, they take care of that. But then when you need to build your own proprietary stack, I see our spend going up and tech. There are still unanswered questions and strategies people can take about SaaS versus proprietary. Then how long are these SaaS tools going to be able to keep an advantage over the underlying models themselves? You see week by week, someone like OpenAI or Cloud, pick an area, legal tech. I'm going to go after that. Here's a new skill that Cloud has. It's fascinating to watch. We're trying to be nimble, stay flexible, not get locked in. These are all the things that we're doing in order to jump from opportunity to opportunity. Is it safe to say that the niche players will be safe for a little while? When I look at the unique access that we have to the documents that flow within this industry to the data sets, OpenAI does not have access to that. Which is why we're super protective about what tools we use and what we lock down. There's plenty of things that are locked down here at Hamilton Lane and things that you're not allowed to use as an employee of Hamilton Lane. If you keep our data within the four walls of these protected tools, you're good to experiment. I look at what we have as still a huge advantage we've got this mode. Because not everything is publicly out there. It's still private by nature of this industry. The winners are going to be those that still have proprietary access to information, as specialty around that. A very thin wrapper around the LLM for a very specific use case. I hate to pick on anyone helping the accounts payable team understand what's happening with invoices. We could probably build a generic agent around, rather than by a special tool that's specialized in that. That might be the place that gets challenged over time. Very specific things like that that don't rely on proprietary data and say, what about APIs? Our goal is to have a two-sided marketplace. I don't know if the liquid markets are the endgame. That's where we're trying to head towards. But you have single-point solutions on the LPSIDE and the GP side. We're slowly finding workflows and quasi-competitors are in co-optition. They're all now partnering with one another to slowly build this workflow. How do you guys think about all those pieces as they come together? We've recognized for years that we can't have our data captured and help at ransom within one of these, our CRM, let's say, our investment book of record. We have to have control over our data. There still are a lot of great uses for those types of tools. There are workflows in our CRM and data capture capability there. We take that data and then we bring it down into our central data fabric. In our data fabric, some of it is highly structured data that's queryable by Power BI tools. Some of it is unstructured data lake type data documents and things that we can reference. We want to have a central hub of data that we can feed into any point solution system that we might want to plug in. On plug, this data can flow directly to cobalt where our clients can access it, whatever it is. But we feel much more confident in the capabilities of analyzing data and mapping things together if we've got the central data hub in place and we're not relying on individual tool to tool API connections. We still have some of those in certain places that might be where you start. Eventually, you got to bring it down. You got to centralize it and you got to master all the data together. Now we're moving into this new world with AI. All of this data needs to get into this new AI layer that didn't exist before. I think about this construct, people come to me and like, hey, should there be a chat interface within this tool or within that tool? I don't think that's where the world is going. That would seem pretty weird to have a different dialog box within every application that you use. I'm already getting annoyed about all the AI features and pop-ups that are happening in PowerPoint just to pick on a behemoth like Microsoft. What I do need is every solution that we have that has valuable data for the firm, every new data vendor that we're talking about, licensing new data from. How is that data going to be connected immediately into this new AI layer that we're building? I don't even know yet what we're going to be capable of building in there, but I do know already you've got the chat feature and you can explore the data sets and I'm seeing really incredible things come up in this generic chat layer that connects all of our systems. That's the new version of where data is going to have to plug in using MCP or whatever protocol it is, but getting it accessible by these agents that we're building that becomes the real unlock. What about the charges? I'm hearing that if you're a SaaS tool and you have an open API and then they're charging their clients access to their own data, which your reaction to that. We look down on that kind of behavior. We are getting some tools that are telling us your AI layer cannot have direct access to our data. You have to call that data on demand instead of having access on the backend to the file or whatever it is that gives you the quickest retrieval. You got to do it in a different way. Some tools are trying to establish a wall to say no, no, we want to keep you in our tool. That's a losing battle over time. People are going to continue to have to deal with this situation where data is going to be free flowing, protected, available, only and permission and that's a tricky thing to handle, but we've got tools that handle it because I can't have access to your email, Scott or your pay or whatever it is. We've got to have said a lot of walls on that. You got to get all this data downstream. We're going to be working with vendors and tools that are making that possible and not setting up friction and not setting up walls. That's a protectionist strategy that I think ultimately will lose out. Be curious to see if there you had working groups behind the buyers and sellers of tech coming together to figure out how do we make this all work? Because I think workflows reporting and all this stuff. There's so much technology coming to bear and the bearity entry is so low. For a lot of LPs, they don't have a resource similar or the size or scale of a Hamilton Lane. How do I get my arms run? All the things that are out there, what's good for me? It's all done with word of mouth. How do I get better at trying to make these decisions and something that is moving really quickly? Internally, we hate committees, but task forces are super important. They are built with a purpose and then you can kill them and let them go. For AI, we built an investment AI task force who brought together bright people from each investment team. We've got a bunch of different verticals. We had a mission. Let's go test these 30 tools. We're getting an email every week from some new tool. It says they're great. Let's give them all the same tests. Let's make them sign the NDA. We give them the test. They get access to this certain data. All right. These 10, what good? Those are the 20 that are out. Now we ask for a pilot, we get access to that pilot, we put them through the ringer, all right. 10 is down to three. Now we talk price and support model, who's going to be a partner to us and who's going to treat us at arm's length like we're nobody. Okay, here's our partner. Now it's one. We went through that in a very systematic way and we've got the receipts to prove our decision. Ultimately, we came out with the best tool, the task force moves on to other things. That's how we did it internally. Certainly, what I rely on is building a network of partners and there are places where we've got frenemies and competitors, if you squand all kinds, but a lot of us on the tech side, view it as valuable to keep those connections up. Is there's only so many vendors that I can talk to, so many things that I can test, so many things that I can stay ahead of. I value having other people's perspective that's encouraged across Hamlet's Lane at all leadership levels. Having a network of peers and other folks that you can talk to, that you can share ideas in a way that's not competitive in nature, but just allows us on a level. Some of the problems we're trying to solve, I need external partners to solve it. We've built consortiums around some of the things that we've done. When each of these deals that I've talked about, canoe and 73 strings, Daphne, I'm working with Apollo, KKR, Blackstone, many groups like that. We do deals with them all the time. We're also chasing some of the same capital. We view it as beneficial to be together in building the infrastructure of the future that tends to help everyone out. That's really cool. There's so much going on in trying to get thought leadership pointing the right direction. I'm pretty sure you hear about things that you have in thought of, different angles and things you agree with and something you might disagree with. I think that's a healthy conversation in the industry. Griff, I'd love to turn to close with two closing questions. The first one is, what advice would you give to an emerging manager from a technological perspective? We say this frequently and many of the thought pieces that we release to put our money where our mouth is. We've invested in groups. I know that a lot of emerging managers don't have the time or spare a capital to go make an investment in an early stage company. You do have the time to invest in yourself and your infrastructure on board technology. You cannot make that the last priority. Sometimes, especially as an emerging GP, any extra dollar of spend comes out of the pockets of the partners. We have seen GPs win over time because they've invested in themselves and their people, but also their tech stack. Progressively as an LP, we are evaluating GPs, how they're set up, what tech they're using, what partnerships they have as part of our operational due diligence. If you fall behind on that, you're going to get a big red mark from us. It's not going to help you raise the next dollars. Do not under-emphasize that aspect of building your organization. The other question I have is what one book article or other resource that you commonly refer to people? There's one book that has come up. It's definitely a classic in the technology space. The innovators dilemma by Clayton Christianson from the late 90s. It's still as relevant today as it's ever been. It's very applicable to the story I've told here about Hamilton Lane. We didn't have that fear of disrupting ourselves. We invested in building our cobalt and digitizing our business. One of the central thesis of the book is that when you build something that's disruptive, that's new and tech. And inevitably, it's not good enough for your best clients. I felt that very discreetly early on. What I built, we initially released the first version of it. It was not meeting the needs of the biggest clients that we had. Very sophisticated pension plans or sovereign wealth funds. It opened up a whole new class of investor, a bunch of family offices that never worked with us before, said, "Oh, this is interesting. I'm going to start a relationship with Hamilton Lane by subscribing to the species tech." And it was good for them. I listened to what they needed and we kept building and we kept building, and sure enough, to turn up along story short. We've been out of for 10 years. Some of my best clients are big pension clients and sovereign wealth funds now. And we ended up meeting them where they needed to be. But it's the classic. We've talked about J-curves in this industry all the time. The book talks about S-curves. The exponential rate of improvement in tech allows you ultimately to deliver that value. They need to do that classic older client base that you had. But you can't be afraid of that early comments, Harry, that what you're building is not good enough because it's not going to be. You're going to meet needs of a different client type. And if you invest in that, good things can happen. Good stuff. Griff, thanks for the time. Look forward to staying in touch. Absolutely. Thanks, Scott. Thanks for listening to the show. If you like what you heard, hop on our website at capitaliselocators.com where you can access past shows, join our mailing list, and sign up for premium content. Have a good one, and see you next time.