In this interview, Ethan Lockshin talks with Rahul Rekhi, President of Rogo, about the intersection of finance, regulation, and frontier AI. Rekhi explains that the biggest mindset shift moving from traditional finance roles to leading an AI startup is recognizing that trust—whether with governments, banks, or clients—remains the foundational element across both worlds. He outlines three pillars that protect Rogo from short‑term growth traps: (1) investors and strategic partners whose interests are aligned for long‑term value, (2) deep client empathy derived from a senior team with extensive banking experience, and (3) a broad AI‑for‑Wall‑Street opportunity set that encourages a multi‑year product roadmap. Addressing concerns that AI might replace junior analysts, Rekhi argues AI will automate drudgery, allowing analysts to focus on strategic insight, drawing a parallel to the historic shift from manual calculations to spreadsheets. He highlights a non‑obvious trend—reinforcement learning (RL) applied to finance—where foundation models are fine‑tuned on elite deal‑maker data to create institution‑specific, high‑integrity financial intelligence. When comparing large language models (LLMs) to traditional quantitative models, he notes LLMs excel at ingesting and synthesizing unstructured, multimodal data and, when combined with RL, can capture nuanced judgment that is hard to encode programmatically. Finally, Rekhi describes Rogo’s competitive advantage as three pillars: cutting‑edge AI models (including RL expertise), deep partnerships with major financial institutions that provide advisory support, and a product/content layer that integrates AI into domain‑specific workflows and data sources, delivering a “last‑mile” solution for Wall Street professionals. The interview concludes with a prompt for Rekhi to share Rogo’s founding story.
Chapters
Given your background in investment banking, government and financial policy, and now you're leading a fast moving AI company, the transition is dramatic. What was the single biggest mindset shift that you had to make when you moved from a traditional finance regulatory role to now being the president of Rogo, a frontier AI company?
Ethan, here's what I think will surprise people. These environments are more similar than they are different. When, whether one is an executive at Rogo or an advisor in government, or an investment banker, each of those roles is predicated on trust. Trust with, for example, senior officials in governments, our government, overseas governments, trust with stakeholders, and in our case at Rogo, with senior banking executives and private equity executives that are adopting the product, and trust with clients. Again, whether those clients are as an investment banker where you're advising them on deals and transactions, or again in the Rogo context where you're advising them on their AI or digital services transformation. So I would argue that there's quite a lot that carries over. Of course, when you're at a startup, you move quickly and you bring a pace and intensity and cadence that the nimbleness of a lean organization like Rogo can capitalize on. But at the end of the day, it's still primarily about cultivating trust and delivering on it.
Rogo Series A was a little over $18 million. You recently raised a series B, $50 million. That means pressure to deliver growth quickly. With those expectations, how do you avoid failing into the trap of short term growth at the expense of long term strategic moat creation?
Yeah, it's a fabulous question. And to me it rests on that principle, rests on three pillars. Pillar number one is having investors that are aligned with that long term growth orientation. And we're blessed at Rogo to have a cap table that reflects exactly that. That's true on our venture investor partners list with firms like Thrive Capital, Sequoia, Tiger Global and others that have built their reputations on being long term partners with companies often taking that, you know, having an ownership stake that runs from very early stage company all the way to a big public company. But that's also true of our strategic partners. You think an entity like JP Morgan, for example, which is invested in us, they're also a customer, a partner, a client of us. And so when we're around the table with them, they're thinking about, you know, again, how we can deliver lots of value for JP Morgan over time and then of course, how that drives Rogo's own performance. So complete alignment around the investor pool is a big one pillar one, pillar two is about client empathy. And so we're also blessed at Rogo insofar as about half our team, including most of our senior team, and in this category as well, we've all worked at investment banks and other leading financial institutions. And so we feel deeply the utility of the AI tools that Rogo brings to these institutions and also what can be built over time as the tools get more powerful, as we get more deeply embedded. And again, because we're focused on that vision as folks who've been on the other side of the table, been users of software, pre AI, that again built in empathy, I think naturally nudges us to be longer term in our partnership orientation with the firms we work with. And then for me, the third pillar is intrinsic in some ways to the opportunity set around AI for Wall Street, AI for financial markets. It's such a rich domain for bringing again AI transformation to work. Everything from modeling to pitchbook and slide creation to ultimately adjudicating transactions agentically. And so because we have that scope and canvas to build toward in terms of our product roadmap, we naturally again are thinking about what we can achieve in five or 10 years and not quick wins based on a more narrow product scope.
There's a growing debate, will AI augment human analysts or replace them? Some fear job disruption. How do you at Rogo address concerns from junior analysts or other finance professionals who may feel threatened by your AI analyst vision?
Right, so the first thing I'll say, Ethan, is coming back to my comment earlier about client empathy. I started on Wall street as a junior associate, so I know full well the nature of that associate experience. And it's my conviction that I will create more work and frankly more rewarding and enriching work for associates and analysts and other professionals of Wal street is rooted in my own experiences. And in particular, I found being an associate immensely fascinating as I was able to move past the drudgery. The kind of manual labor of making sure the right logo is in the right place and data entry was conducted accurately and more toward helping a client make an ultimately transformative decision for their company, whether that's an acquisition or an IPO or something specific similar, that was immensely stimulating. But you had to trudge through that very in the weeds grind to get to that point. And I think one of the reasons we're all excited about Rogo and AI in this space is it allows talented professionals to get right to that point of delivering strategic insight to a client and allow the machines allow AI to traverse that basic drudgery. Do the groundwork, do the legwork so that you can actually deliver great advice. And that's what we're excited about first and foremost. I will also say that this is in many ways a natural extension of a long running debate and threat on Wall Street. You know, our predecessors across the generations, for example, and the FT has done some great reporting on this, were focused on when electronic spreadsheets just came out and we had VisiCalc and the first versions of Excel. There was a concern of what's this going to do to all our junior professionals that are manually doing calculations in physical versions of spreadsheets, when again you can have software that does these calculations instantaneously. And if you look at where we are now, we have more finance and investment banking and private equity etc. Professionals than we've ever had in the history of financial markets. And the work they're doing is inherently more intellectually stimulating, enriching and value add than the work of again sort of manually doing calculations on paper pre visicalc. And so again, lots of conviction that the same will be true of this generation of technology and that will ultimately open up new vistas and markets for the kind of great financial advice and transactions to be delivered because of the efficiencies and the economies of AI being able to do that base layer of work more or less autonomous. So I think it's an exciting future for finance professionals.
What non obvious AI trends are you paying attention to that most people in finance are missing?
A big one and I should say at Rogo, the beneficiary of being side by side with just some of the best AI researchers and talent in the world. So I have a sense of what's coming down the pike out of the expert communities around that topic. The big one coming out of that thread is the power of reinforcement learning in finance. Reinforcement learning is a, is a paradigm for building AI models that involves taking as an input the foundation models that come out of places like OpenAI and Anthropic and Google Gemini and so forth and you know, in effect training them on what good looks like to elite human professionals in e.g. high finance. As it relates to again building financial models, delivering great advice around a transaction for a client, having fairness opinion materials, for example, that are incredibly high integrity and that can withstand the scrutiny of the public markets for every facet of what makes a phenomenal banker or investor, building a model that knows what good looks like and uses that to train that general financial, the general intelligence model so that it becomes an elite thinker and Doer as it relates to finance outputs. And what reinforcement learning rests on is having that number one sort of representation of, you know, these are, you know, 10, 15 years of excellent pitch decks, S1 filings in an IPO context, merger models, et cetera, from the leading, for example, investment banks in the world, you know, many of whom are our customers, and have that, you know, again, feed back into the reinforcement learning paradigm. So the model learns what good looks like without us explicitly telling the model to do a certain thing or to do another thing. And second, deep embedding with the sort of internal context, processes, judgment of elite financial institutions. So again, these models can learn from the best deal makers that they have. We have the privilege at Rogo because of our position in this AI for Wall street ecosystem to be building toward that vision of, in effect, whether it's a JP Morgan or a Lazard or what have you, they're not just getting the Rogo financial intelligence that any firm can show up and purchase seats for and so forth, but a version that's using reinforcement learning trained on their best dealmakers, that's very bespoke to that institution itself. And we think the power of that over the next 12, 18 months is going to be completely transformative above and beyond what the commercial AI that all of us are using is already doing. So that's a trend that I think is on top of mind for the AI community. A little under the radar right now in the world of Wall street and finance, but it's going to be top of mind for everyone within 12 to 18 months.
What do you think large language models enable that traditional quantitative models never could?
It's a great question. So I kind of put it into two categories. So category number one is dealing with large volumes of unstructured data across formats. So multimodal as well. When I was at Lazard, Lazard has still to this day, but had at the time set up an excellent data science team that did that kind of quantitative analytical work on behalf of investment bankers for client situations where they would, for example, pull in for a consumer client data on different retail locations of a retailer and then do analytics around what share of different slices of America or Europe or what have you live within certain radii of those retailers, etc. That kind of structured analytics they were great at, they continue to be great at. And that's the kind of thing that traditional machine learning techniques, you know, again, have, have had great performance at delivering. Where LLMs really thrive is in wider sets of unstructured data across text from filings, for example, or research reports, hard structured data like tables, for example, within filings or within press releases, transactions, data from databases, etc. They can traverse across all of these and synthesize insights or summaries or patterns in a way that's coherent and flexible. So to me that's, that's one big pillar. The other pillar is coming back to our topic about reinforcement learning. If you show an LLM what good looks like in the right reinforcement regime, reinforcement learning regime, it will naturally create connections and distill learning that is very difficult to instill programmatically in the context of traditional machine learning techniques. And so if we, we stop and think about what makes a great a banker, for example, there's so many different rules, contingencies, heuristics that one could spend a lifetime writing those down and still not really capture the nuances of judgment. But the power of LLMs and agents is that with the right reinforcement learning, in the same way that you can train AI to beat the world's greatest go players, for example, you can train them to think like the world's best deal makers. And that's again very specific to this regime and this paradigm.
You just raised 50 million Series B Rogo, it's one of the fastest growing companies in the AI space in the finance sector. What is Rogo's competitive advantage?
Yeah, I would put it in three categories, Ethan, are models and AI number one, our partnerships number two, and our product and content number three. So let me just take each of those in turn on models in AI, we have some of the world class experts in AI regimes like reinforcement learning that give us an edge, that give us real differentiation when it comes to standing on the shoulders of the excellent work coming out of the foundation model labs, taking the best at what each foundation model lab has to offer for different workflows. So for example, as OpenAI and Anthropic and Google, Gemini and others continue to invest in spreadsheets and slides and chat and research etc. Being able to curate the best of each one for different types of work that financial professionals produce and then layering on top of that our own reinforcement, fine tuning agent, orchestration and other innovations so that our partners, our customers, are getting the best of what AI has at any given point in time for work of Wall street professionals, that's piece number one. Piece number two is, as I mentioned, partnership. So when we work with large financial institutions, it's not just sending people a login link to a product and walking away, but having large teams of, we call them advisory professionals, but they're in effect kind of forward deployed bankers and investors, people who've often worked at these institutions that are our customers, in effect enable the bankers, investors, et cetera, that use Rogo to get the most out of the platform by working with someone at Rogo who's been in their shoes, speaks their language, understands their use cases, and can help them get over the skill curve as it relates to AI and maximizing its use in their work. And then the third piece is the product and the content to really transform Wall Street. It's not just about unleashing AI on a piece of work and letting people ask questions, et cetera, but pairing the power of those conversational AI tools with domain specific workflows, whether it's screening or model building or integration with data sets from financial data providers like the London Stock Exchange Group S and P, so that it all lives natively within the single pane of glass that Wall street professionals engage with to do their work. And that, you know, you can call it the last mile, but it's almost the last hundred miles for how you go from again, kind of general AI to something that really solves the problem for Wall Street. And that for us, those three pillars together is really what Rogo is all about.
Raoul, you recently joined Rogo as the President. Rogo is one of the fastest growing AI companies in the finance space. You just raised a $50 million Series B. Tell me about Rogo and the founding story.
Yeah, well, and I remember it fondly, Ethan, because one of our founders and our CEO, Gabe Stengel, worked with me as an analyst at Lazard. We were both in the Healthcare group. Gabe had a dual role across the Healthcare group at Lazard and the Data Science group at Lazar. And much of his early insights behind Rogo were premised in part on seeing both sides of the house. Seeing, for example, the power of what bringing machine learning into the work of investment bankers could do, how it could, for example, elevate junior bankers in their work beyond the day to day manual kind of drudgery and allow them to really step up and be analytical partners and thought partners to clients. But also from the other end, coming at it from a banker seeing what it would take to close that last hundred miles of machine learning innovation to actually transforming the work of bankers. Early on, I remember vividly Gabe sketching out, you know, almost the proverbial cocktail napkin idea behind Rogo and talking about, hey, should I leave my incredibly prestigious rewarding job at Lazard to go chase this dream. And I'm I guess, fortunate to have given him the advice, I'm sure, along with others, to go pursue it. And of course, Rogo became what it is today. I think Gabe always had this vision and this was even pre chatgpt of what machine learning could do in banking. And then the release of ChatGPT and contemporary LLMs really turbocharged the business. And that kind of connects to how I joined, which is I was an appointee at the U.S. treasury. I had followed Gabe's story, of course. We were friends from our time at Lazard together and seen all the progress he had been making. When I left the US treasury, as my appointment came to a close with the transition, started thinking about next steps and Gabe and I caught up as part of that. And he said, you know, we were, we were catching up in Rogo's offices. He said, you know, Raul, I think you'd be blown away by what's coming down, coming down the pike with AI. And I said, look, Dave, there's, there's a lot of hype around AI and tools, you know, make it tangible for me. Let me, let me sort of see some of your early products and what's on the roadmap. And I spent a couple weeks with what became, you know, we called it Rogo Pro, but it's sort of a full agentic reasoning, model based orchestration layer for Wall street work. And I gave it the kind of questions and work that I had been addressing as a banker just before I left for treasury. And I was blown away in that it replicated 98% of what I was telling a board at a Fortune 500 company to do, to navigate a very delicate activist situation that to me really lit the fuse of what Rogo could become and of course, conviction in the team and the product. And I couldn't be more thrilled to be part of that team today.
