The AI-Powered GP, Part Two: Building an Operating Model That Enables Growth

AI can help GPs build more scalable operations, leverage institutional knowledge and stay focused on delivering returns.

As private equity (PE) firms grow and become more complex, can they continue relying on the same operating models they used 10 or 15 years ago?

Dominic Dalli: Our firm was founded 25 years ago. In fact, this year we're celebrating our 25-year anniversary. Like a lot of PE managers, we very much started as a founder-run firm with a single strategy.

The biggest change for me has been the fact that we've had to institutionalize a lot of our processes. We've had to, if you like, release the business from the original founders' influence, and that has required a significant level of investment in infrastructure.

The kind of people we're employing within our firm these days are very different and have a range of skills different from the founders of our firm 25 years ago. Introducing a range of tools to manage multiple funds safely, communicate with our LPs effectively, develop our network and develop our business and strategy accordingly has been absolutely fundamental, particularly in the last seven or eight years.

What are the warning signs that a firm's growth is beginning to outstrip its operating model?

Olivier Mangez: When deal flow is slow, you don't really see those signs. It's when the number of deals starts to increase or you're launching your fundraising process that you start seeing those signals.

I see two different signals that are key. The first is the human bottleneck. That's when everything in your organization relies on senior members of the team who have the memory of the company and the deal flow, and sometimes the process slows down because they're not available. That's something critical that you really need to consider.

The other aspect is that the fundraising process is a stress test for the organization. Will you be able to respond quickly and efficiently to LP requests? This is where information management is really key to being able to share knowledge across team members, even new joiners.

I think something that's changed drastically in the market over the last few years is the emergence of generative AI, which can help manage this flow of information that can sometimes be overwhelming.

As managers take on more complex fund structures, what role can technology play in helping them scale?

Dalli: Like a lot of firms, we went through that step change of investment in technology many years ago. I think it's a bit of a hygiene factor, frankly. As always, it's best to invest ahead of the curve and make sure you've built the infrastructure before you go and launch a bunch of new products.

Then it's about scaling that technology. Yes, adding people to a degree, and we've done that. We have a very meaningful and significant finance, operations and compliance team, and they're absolutely embedded in our business. They're really critical to what we do.

But we're supporting those people with what is now a range of service providers and products. That might be around NDA management, cap table management or portfolio monitoring, for example. I'm always surprised by how many of these product providers exist and how innovative they are in offering solutions to problems that, frankly, we never really knew existed.

What can AI automate during a live deal process, and how do firms balance that with human judgment?

Mangez: Automation is great. It should not replace judgment. We need to have a human in the loop. Technology is an enabler. We want to keep our people making the critical decisions.

From a technology perspective, what we have available today is really incredible. When it comes to dealmaking, there are some important considerations. Decision-makers ultimately want insights, so are you feeding the technology the right information for it to give you those insights? That's absolutely key.

It's very easy to say, "I'm connected with ChatGPT" or "I'm connected with Claude." It doesn't mean that you're feeding the AI the right information.

The second thing is that Claude and ChatGPT are very good at compressing information. They'll take information from a lot of different sources and merge it into a summary. Amazing. The thing is, if you merge a lot of information that some users shouldn't have access to, once it's bundled, it's very difficult to extract the confidential information.

From a technology perspective, what's absolutely key is deploying technology that's able to hide information from users who aren't supposed to have access to it.

Especially when you're exposing data to a buyer, you don't want to expose all the information. So, I think there are two critical aspects: Do you feed the right information to the AI, and do you control what is made available?

Have you used AI in your due diligence process, and where have you found it most useful?

Dalli: I think we, along with a bunch of other lower mid-market, mid-market and bulge-bracket investors absolutely use a level of automation for company monitoring, investment reports and internal investment committee submissions.

I'd say the quality of those reports has improved, particularly over the last couple of years. You can find that with relatively little human input, the quality of output can be really quite good.

The game changer for us around the deal process has been deal origination and what automation can do in deal origination.

We track 16,000 companies within the sectors that we invest in that are of the right scale and founder-owned to be investable for us. Many years ago, we would have tracked far fewer companies, and we would have been very reliant upon the knowledge of our deal originators as to how those companies were progressing and performing.

Now, as you'd expect, we have a very sophisticated CRM system that updates itself quarterly and is updated by our deal originators quarterly on company news flow and financial performance.

Therein, for us, lies the answer to your question. It tends to come down to what type of due diligence we're looking to do. Are we looking to track those businesses that are punching above their peers in the niches we're most interested in informally before we engage in formal financial due diligence?

For that aspect of the process, I think automation is super powerful, and we've found it of increasing use over the last few years for sure.

As automation and AI become more common within deal teams, which skills become less relevant and which become more valuable?

Dalli: The obvious one for us is financial modeling. There's no doubt that automation can play a very important and, frankly, reliable role in developing pretty sophisticated financial models.

Clearly, it's important that our deal teams understand how these models work, but they're absolutely able to generate these models through the support of automation so much quicker than they used to.

The bit automation can't replace is persuading the founder-owner to engage in a PE deal for the first time in his or her lifetime. Therein perhaps lies the answer to the importance of the human in this whole dealmaking process. So absolutely, it's useful.

Olivier, which skills do you see becoming more important as firms adopt AI?

Mangez: We need to adapt. In every organization, it's a new technology. Everybody needs to learn. There is no user manual for it.

You need to learn what it's capable of. You need to confront your assumptions with reality. I think the most important aspect is curiosity. You need to embrace technology. You try it and see what it's capable of.

You realize what it's not capable of — very important — and you experiment.

What are the operational costs of sticking with old-school, manual processes as a firm becomes more complex?

Dalli: One of the beauties of technology is that it's pretty cost-effective. If I were to compare two major macro dynamics over the last 20 years — the adoption of technology versus the increase in regulation — I'd say the costs associated with increased regulation in running a multi-product, much more complex private equity manager are multiple times more expensive than the introduction of technology.

From our experience, the introduction of technology has been a real enabler. It's brought a lot of efficiencies.

We've supplemented that, as I said earlier, by increasing our compliance, operational and back-office teams, for sure. But I don't think it's been cost-prohibitive for us. In fact, I think it's been very cost-effective.

How can a consolidated technology environment help firms maintain standards, control permissions and create an auditable trail?

Mangez: Standardization only matters if it survives pressure, right? The pressure of deals. You have a deal coming onto the table, you have time constraints and resource constraints. You have a fundraising process being launched and inbound queries from LPs. This is where you test whether your standard actually matters or not.

If for every deal you're making, every add-on and every bid, you're pressing the reset button, what's the point? You're not really learning.

Introducing learning practices where you learn from every deal you're making, consolidate those practices, try to automate them and introduce a feedback loop where you improve really matters.

When you consolidate your tech stack, you can create your checklist of all the questions you know will come from the investment committee. What are all the questions or topics that LPs may inquire about during the fundraising process?

When you have the right technology and the right information management system, you can standardize practices and processes to make sure that information is ready the day you need it.

I think this is what matters when you invest in technology: making sure you can repeat processes and improve processes over time.

As the industry grows and operations become more complex, how do you maintain oversight and visibility across the business?

Dalli: We have the benefit of being a single-office, single-country, single-strategy manager. Being present is really important. Our teams are co-located here at 25 Victoria Street, and through daily interaction with my teams, I can get a good sense and a good handle on what's going on, where certain functions are struggling and where there's innovation that we can learn from within the organization.

For an organization such as ours, keeping oversight is relatively straightforward.

I think more challenging for us is the ability to consistently innovate, but within our fairway. It's about reminding ourselves why we have the right to exist and why we've been able to generate alpha for our LPs for so long.

While they want us to be innovative and come up with different products that satisfy their needs, we've got to remember what we're good at and really what our right to exist is.

How can technology and AI help larger organizations with multiple strategies or geographies maintain oversight and control?

Mangez: I think one of the big risks in an organization is not having a gap and knowing that you have a gap — it's completely missing the fact that you have a gap in your knowledge because you're so overwhelmed with information that you don't really see it.

This is where I feel technology can help us. As I mentioned earlier, the real benefit of AI is its ability to compress information, make it easy for a human being to ingest and help spot inconsistencies and gaps when we're evaluating a deal, working on a fundraising process and so on.

It's about seeing the big picture. Now we have tools. Five years ago, we had no tools to do it. It was really relying on your brain.

Putting technology aside, what other challenges come with scaling a private equity firm?

Dalli: Ultimately, we exist because we generate outsized returns. As private equity firms scale, I think it's really important that they don't lose sight of the fact that maintaining and improving returns is their reason for existing.

A second aspect is culture. As our firm and other firms institutionalize, bring in a different range of skills and invest in technology, the culture of those firms can really change.

Again, we mustn't lose sight of the fact that we want to be ambitious, we want to incentivize, we want to retain and we want to motivate our people. I see that cultural development as being equally important.

Is the gap between the back office and front office closing, with operations increasingly becoming an enabler of investment success?

Dalli: The phrases back office, middle office and front office, for me, always sound like investment banking phrases rather than private equity. But I think that's true for a reason.

A lot of PE firms have been very good at making sure that all of their colleagues and everything they do are absolutely integrated into their deal process.

Fundamentally, we are here to serve our LPs. If our LPs are looking for us to diversify strategy or are looking to invest in, let's say, co-investment opportunities or continuation vehicle opportunities with us, we have to listen to that.

We have to think about where within our deal process we can offer our LPs what they're looking for. I think private equity has been fantastically innovative in designing a range of products for its LPs, making sure that the primary capital — the flagship fund — is continually fed by being able to offer alternative sources of investment to its LPs.

Has a rigorous operational model now become a hygiene factor for LPs when deciding which managers to back?

Mangez: Something really important here is governance. You have two ways to see it: Either you see it as a constraint or you see it as an enabler. If it helps you grow, then it's not back office anymore. It becomes front office.

How can you execute your deal faster? How can you execute your deal more reliably?

We were talking about LPs. They expect transparency. They expect data to be available. A few years ago, providing transparency to your LPs was a unique selling point. It was a differentiator. It's not anymore. If you don't have it, you're filtered out of the list. So it's really key.

But something to keep in mind is that the best firms don't really talk about technology. It's about what the technology allows you to do. Can the technology help you be more reliable, repeatable and resilient? This is what matters.

Technology is not the ultimate goal. Technology is the enabler. This is what we need to keep focused on.

Dalli: Indeed. You made a great point earlier about the fact that we need to make life easy for our LPs, whether that be around transparency or due diligence, for example.

For me, it's fascinating that our industry has grown exponentially, as you pointed out at the start, Nick, but I'm not convinced that the due diligence capabilities and due diligence teams within our LPs have grown at the same rate.

Therefore, we need to make life easy for them.

Mangez: One hundred percent.

Ready to scale smarter? Learn how FundCentre AI and DealCentre AI can help your team streamline workflows and turn fund and deal data into a competitive advantage.