Alt-X Is Building Cursor for Excel, and Wall Street Might Actually Use It

The Macro: Private Markets Still Live in Spreadsheets

There is a meme in tech that spreadsheets are dying. That AI will replace Excel. That we are entering some post-spreadsheet era. This meme is wrong, and anyone who has spent time around private equity or real estate investing knows why.

Private market deals live in Excel. Not because people love Excel, but because the models are too complex and too customized for anything else. A single real estate underwriting model might have 50 tabs, thousands of formulas, and assumptions that change with every deal. These models are built from scratch or from templates that are barely better than scratch. Analysts copy numbers from PDFs, CIMs, and data rooms into cells by hand. They rebuild the same structures over and over.

Every year, trillions of dollars move through these spreadsheets. The models are fragile. One wrong cell reference can cascade through the entire analysis. And because every firm has their own approach, standardized software solutions keep failing. Argus does well in commercial real estate. But for private equity fund modeling, leveraged buyout analysis, and complex waterfall calculations, nothing has replaced Excel.

The opportunity is not to kill Excel. It is to make Excel smarter from the inside.

The Micro: A Stanford Quant and a Cornell Engineer

Ryan Samadi and Michael Wachsman founded Alt-X. Ryan studied CS and AI at Stanford and previously worked as a commodities trader at Citadel, where he led a hedge fund investing program. Michael studied CS at Cornell and focuses on infrastructure engineering. Garry Tan is their YC partner, which tells you something about how seriously this is being taken.

Alt-X is building an Excel agent that reads deal documents, understands the deal structure, and builds or populates financial models automatically. The pitch is straightforward: instead of an analyst spending 40 hours copying numbers from a PDF into a spreadsheet and building formulas from scratch, Alt-X does it in minutes.

They already have hundreds of millions in AUM on the platform, and users reportedly prefer it to legacy cash flow modeling tools. That is meaningful traction. Financial professionals are deeply conservative about their tools. If they are actually switching to Alt-X, the product is delivering real value.

The key differentiator is trust. Financial models cannot have errors. A missed decimal point can mean millions of dollars in mispriced risk. Alt-X needs to be not just fast but accurate in a domain where accuracy is existential. They are positioning themselves as “the Cursor for financial modeling,” and that analogy works. Cursor did not replace code editors. It made them smarter. Alt-X is not replacing Excel. It is making Excel smarter for a specific, high-value use case.

The Verdict

Alt-X is attacking one of the most valuable software problems in finance. The total addressable market is genuinely enormous. Every PE firm, every real estate investment shop, every family office that underwrites deals manually is a potential customer. And the switching cost from “doing it by hand in Excel” to “doing it by AI in Excel” is nearly zero, because the output format is the same.

The risk is competition from established players. CoStar, Argus, and newer tools like Coyote Software all target parts of this workflow. If Cursor itself adds financial modeling capabilities, that could be a threat. But the deep domain specificity of private market models is a real moat. You cannot build a good financial modeling agent without understanding waterfall calculations, cap rate analysis, and the specific ways that different asset classes are underwritten.

In 30 days, I want to see error rates. What percentage of the time does Alt-X produce a model that an analyst accepts without correction? In 60 days, the question is deal volume. How many actual transactions have been underwritten using Alt-X models? In 90 days, I want to hear about firm-wide deployments. Individual analysts adopting a tool is interesting. An entire PE firm standardizing on it is a completely different signal.