Gabriel Stengel on Building the AI Platform for Investment Banking

Gabriel Stengel knows the junior banker grind from the inside. He worked long hours as an investment banking analyst at Lazard before founding Rogo, an AI platform built specifically for investment banking and private markets. Today Rogo counts Jefferies among its users, works with the majority of bulge bracket banks, and sits at an unusual intersection: it is simultaneously a customer, a distribution partner, and a potential competitive target of OpenAI and Anthropic.

At the Jefferies Private Growth Conference in Santa Monica, Stengel described the opportunity to meet other founders who are “a step ahead of me, more mature, who have seen how to build a business” as genuinely impactful. Stengel sat down to explain how Rogo is building a durable position in one of the most skeptical and relationship-driven industries in the world.

Investment banks are notoriously resistant to new technology. How did you get traction?

Investment banks are skeptical of new technology, but they’re also ferociously ambitious. When there’s clearly technology that makes them smarter and more efficient, they want to be first adopters. The early days of Rogo were built on the empathy I had for the junior banker role: the material creation, the grind, the slog of that work. Being able to speak directly to that experience resonated with the market in a way that a more generic pitch never would have.

There’s a comparison to be made with Bloomberg. Do you see Rogo becoming the Bloomberg of AI for finance?

There’s no doubt there’s an opportunity for a hundred-billion-dollar company that reinvents how AI works in finance. But Bloomberg’s secret is that it’s a massive business built primarily around public markets data. Most M&A bankers don’t actually use Bloomberg. Most sponsors don’t use Bloomberg. What Bloomberg did for public markets, creating data analytics and communication infrastructure, can be done for private markets: bringing capital markets infrastructure into M&A, into the long tail of sponsor transactions. Our strategy is to start with an agentic platform that has financial data and financial workflows, get it into firms, make them smarter and more efficient, and then build out the infrastructure around deal coordination, deal sourcing, and diligence. The goal is to make the entire M&A process more efficient.

OpenAI and Anthropic are both moving into financial workflows. How do you think about competing with the infrastructure you’re building on top of?

The dynamic of horizontal infrastructure providers also wanting to play at the application layer isn’t entirely new. Think about how AWS and Azure competed with tools like Snowflake while also offering overlapping services. For us, we believe there is so much product surface area beyond just the intelligence layer. If OpenAI or Anthropic are targeting the Microsoft Office equivalent for finance, creating better PowerPoints and Excel models, that’s one piece of it. But your best bankers at Jefferies aren’t the ones creating materials all day. They’re out building relationships, negotiating, meeting clients. There’s so much more that goes into coordinating a process. Over the next 18 months, as intelligence gets commoditised, the firms that win will be the ones that deeply integrate across the full deal lifecycle.

The most valuable data an investment bank holds is its own proprietary deal history. How do you unlock that safely?

This is genuinely one of the hardest problems to solve. Every SIM, every model, every email, every management presentation sitting on a bank’s servers is the holy grail of institutional knowledge. It’s also the most heavily regulated data in financial services: Chinese walls, MNPI controls, role-based access, jurisdictional restrictions. We’ve been working with institutions like Jefferies and the majority of bulge bracket banks to help them deploy AI on top of live real deal data instead of generic workflows. That requires building compliance infrastructure together with the client: monitoring every piece of data that flows in and out of the platform, how it gets stored, and what oversight sits on top of it. It’s never a linear path to a fully automated system. It takes sustained partnership to get there.

What financial workflows will AI change most fundamentally over the next three to five years?

There are around 300,000 American businesses generating $3.5 trillion in revenue that have never spoken to a Wall Street investment banker. As it gets easier and cheaper to go out and win business, institutions like Jefferies will be able to offer products to that long tail of underserved companies, both in the US and in emerging markets. The second-order effects of financial intelligence becoming dramatically cheaper are going to be profound. Products and services that were previously completely impossible will become viable. That’s the more interesting long-term question: not just making existing workflows more efficient, but what becomes possible that wasn’t before.