Ori Eldarov
Runs the AI Bank.

Ori Eldarov is trying to run a Goldman Sachs process on a plumbing distributor. Not literally the same process - a Goldman process on a $30M-revenue plumbing distributor would cost more than the plumbing distributor - but the same shape. The teaser, the CIM, the buyer list, the auction, the negotiated deal. All of it, but with software doing the analyst work, so the economics land somewhere on the correct side of the transaction.

This is what OffDeal does. Eldarov co-founded it in 2024 with Alston Lin, an ex-Meta engineer, and put the firm through Y Combinator. In July 2025, ten months after its seed, OffDeal announced a $12M Series A led by Radical Ventures, with Y Combinator, Rebel Fund and Centre Street Partners along for the ride. The tape puts the round at a roughly $100M valuation. The company is 33 people. Thirty-plus sell-side mandates have gone out the door in under a year.

The Positioning Fight

The thing to notice about OffDeal is what Eldarov refuses to call it. He does not call it a fintech marketplace. He does not call it software for banks. He does not call it a platform. He calls it an investment bank. This is a strong claim from a company whose fifth employee joined last winter, and it is on purpose. In a long X post from June 2025 he wrote, in classic banker-turned-founder register: "Last year I raised $5M to build a brand new investment bank from scratch. Not software for banks or a fintech marketplace - a real M&A advisory firm."

The line he uses in conversation is more direct. "We're building what Goldman Sachs would look like if it started today." There is a version of this quote that reads like founder theater and a version that reads like an actual operating hypothesis. Eldarov, six years at RBC Capital Markets in New York and London, roughly $20 billion of advised deal volume attached to his name, appears to be running the second version.

The Cost Curve

Small business M&A is not underserved for lack of demand. It is underserved because the unit economics of a traditional sell-side process do not clear. An analyst at a real bank costs a lot of money per hour. The deliverables an analyst produces - the CIM, the working model, the buyer list, the tracker - take a lot of hours. For a $500M transaction, this is fine. For a $15M transaction, it is malpractice for the bank to take the mandate. So the bank does not take it.

The bet OffDeal is making is that the analyst deliverables can be produced, if not entirely by AI, then in a workflow where the AI does the first eight passes and a human does the ninth. Eldarov puts the point sharply. "Yes, you're spending hundreds or thousands of dollars in compute, but it's still cheaper than banker time." If that sentence is true, and it appears to be, OffDeal can price into a market segment that legacy firms cannot.

The Harvard Detour

Eldarov did an MBA at Harvard Business School. He is unusually honest about why. In a post from June 2025: "I did an MBA at Harvard to have work visa security while I work on my startup. I had a great time, but in practice MBA is more likely to convince someone NOT to be a founder, than the opposite - you're rubbing shoulders with very risk averse people. I have since also done YC." The implied ranking - YC above HBS as founder training - is not the sort of quote a business school likes to see attributed to a recent graduate. It also happens to be the quote most likely to end up on a slide at another business school.

The biographical arc is undergraduate at the University of British Columbia, six years at RBC across New York and London, HBS during the pandemic-plus era, OffDeal on the back end. He is 32. The X handle is @leveredvlad, which contains more information about his sensibility than the LinkedIn tagline.

The Recruiting Post

Eldarov hires the way a founder in 2025 hires, which is by posting. When OffDeal needed bankers this summer, he wrote: "We're drowning in deal flow & need help. If you're a banker interested in running your own deals & using cutting-edge AI, we'd love to hear from you." It is a specific inversion of the normal banking pitch. The traditional VP job at a large bank has ninety percent process management and ten percent client work. Eldarov is offering the reverse, with an implicit promise that the process management is being handled by something that does not sleep.

The Customer

OffDeal's stated sweet spot is companies with $10-100M in revenue and $1-10M in EBITDA. This is the segment that private equity partners refer to, without judgment, as the lower middle market. In practice it is the segment where the seller is often the founder, sometimes the founder's kids, and where the buyer is usually a search fund, an independent sponsor, or a strategic three towns over. The seller has no CFO. The buyer has one lawyer and a spreadsheet. There is no natural investment banking incumbent because no one large enough to be an incumbent will show up for the fee.

OffDeal shows up. The pitch is that a seller in this bracket gets a real teaser, a real CIM, a real curated buyer list and a real auction. The AI writes drafts of the analyst deliverables. Humans - Eldarov, his colleagues - run negotiation and closing. The stated outcome is a competitive process rather than the more common non-process, in which a business broker forwards the listing to a mailing list and hopes.

What's Actually Interesting

The strange specific about OffDeal is not that it uses AI. Everyone uses AI. The strange specific is that Eldarov chose to build a firm that is legally, structurally and reputationally a bank, and to put the AI inside it, rather than the more obvious move of building software and selling it to existing banks. The obvious move is a better story for investors who like SaaS gross margins. The Eldarov move is a better story for customers, because the customer of a sell-side M&A process wants an advisor, not a subscription.

The other strange specific is speed. Seed to Series A in ten months, thirty mandates launched, CNBC appearance in the summer of 2025, a Radical Ventures term sheet led by a fund not typically known for financial services investments. Whatever the eventual outcome, the pacing tells you something about which end of the market moves fastest when the tools change. It is not the Fortune 500 end.

Timeline

Frequently Asked