Why I Joined Hebbia as Principal AI Strategist
The first letter I ever mailed to the United States was addressed to NASA, asking about how they engineered the Pathfinder mission to Mars. I don’t remember whether I received a reply, but I remember the drive behind the question: I wanted to understand not only the what, but also how frontier problems were being solved.
That drive to work at the frontier, where the hard problems get solved, has motivated most of my professional life. The opportunity to shape how AI is embedded into finance represents an evolution of that frontier. High-trust, reliable AI for finance, must start from a deeper understanding of how its professionals actually think and work. Finance workflows are chained: data and intelligence feed into complex analyses, these analyses converge into judgment. Value comes from leveraging the whole chain, not any one link. The focus on this pain point and the drive to solve are what brought me to Hebbia.
A decade in banking, at every level
My decade in investment banking across bulge brackets and boutiques has meant executing on what is required at every level. From the late nights inside documents and models as a junior banker, to making the judgment calls a senior banker owes a client, I have experienced where existing processes work, and where they break.
Banking concentrates an extraordinary amount of intellectual capital, then spends too much of it on low-value work. Junior professionals spend most of their time assembling data, not interrogating it. The value a banker creates lies in analysis and judgment, and too few hours are spent here.
A similar misallocation runs at the senior end. A senior banker's craft is synthesis. They draw from a career of complex deal experience and turn a flood of information into context and a judgment call a client can act on. However, one senior banker's memory is only a fraction of what the firm holds. That includes, for example, every comparable transaction the firm has touched, and the lesson inside each one. For most of my career, reaching that full record at the speed a live deal demands has been nearly impossible. That recognition sent me looking outside banking for an answer, and I did not have to look too far.
Even at the earliest iterations of today’s AI models, I was already following the field closely. Around 2023 and 2024, these models crossed the threshold from novelty to genuinely useful. What held my attention was not where the technology stood on any given day. It was the slope of progress. Anything improving that fast was going to reshape the work I had spent a decade learning. I needed to be on the building side of that change.
What I found at Hebbia
Hebbia operates at the frontier of vertical AI, set on changing how investment banking professionals think, work, and deliver advice to their clients. When the call from Hebbia came, I did not take the premise on faith.
I spent time on the platform. With Matrix, I saw a product that respects the complexity of client workflows and addresses the problem I had been circling. For the junior banker, Matrix turns a routine analysis into something built once and run repeatedly: the diligence pass, the comp set, the data-room review no longer begin from a standing start. For the senior banker, it can bring the institution's own record within reach and at the speed a live deal demands. How much of that record Matrix reaches scales with how deeply a firm integrates, and Hebbia has built the technical and compliance infrastructure to match. It gives the professional room to think, not just produce. Matrix returns depth to those who bring depth to it and does not flatten a complex problem into a quick answer. Hebbia is built by people who understand this distinction. It is no accident of engineering and reflects who is doing the building.
Hebbia runs deep on talent. Danny Wheller charts how it wins the market, Barry Duong translates its technology for practitioners in the field and Aabhas Sharma shapes the product as President and CTO. The same conviction that ran through all of them surfaced most directly in my conversation with George Sivulka, who at the end asked me, "Why work hard?"
I think about that question often. The first-level answer to almost any question is easy to reach, and it is often uninteresting. The second- and third-level answers are where the value lives. They have to be earned with hard work. That is true of a hard problem in mathematics, and it is true of the advice a banker owes a client. A founder who asks that question, and means it, is someone interested in far more than technology. That was the quality I was looking for in a leader.
I joined Hebbia because rarely do experience, interest, and opportunity converge. The frontier does not stay still, and I have wanted to be near it since that letter to NASA. What sits at the edge of vertical AI today will look ordinary in a few years, and the hard work will have moved elsewhere. I came to build at that edge. I plan to stay there as it moves.
