A former Goldman Sachs analyst has raised $22 million for a new AI venture, Bloomberg reported, and you should read that as more than another seed round. Wall Street's juniors are no longer only feeding the machine. Some of them want to own it.
Bloomberg's deals desk, in a June 30 report from Manuel Baigorri, said the founder had worked as an analyst at Goldman Sachs before raising $22 million for an artificial intelligence company. The report did not give enough public detail to build a neat founder profile, so don't pretend the name is the story. The useful point is sharper than that: the people who spent their early careers living inside Excel models are now being paid to turn that work into software.
Start with Mo Achour. According to eFinancialCareers, he spent eight years in mergers and acquisitions at Goldman Sachs in London before founding Axe AI, a startup built around producing institutional-grade financial models in minutes instead of weeks. Achour has said he reviewed and built thousands of models. That detail matters. You don't need a founder who has read a memo about banking workflows. You need someone who knows what a managing director sends back at 1 a.m. because the debt schedule is wrong.
Axe AI is not the same company as the Bloomberg deal, based on what is publicly known. It is the useful comparison. The pitch is not vague automation for finance. It is a banker saying: I know this work well enough to compress it.
OpenAI has seen the same opportunity from the other side. Bloomberg reported in October that the company had hired more than 100 former investment bankers from Goldman Sachs, JPMorgan Chase and Morgan Stanley for an internal effort known as Project Mercury. The contractors were reportedly paid $150 an hour to help train systems on financial modeling, including the kind of spreadsheet work used in IPOs, restructurings and other transactions. Applicants had to pass an AI-run interview and then submit a model every week.
Look at the split. One former banker takes the hourly rate and helps a larger AI company train its model. Another raises venture money and tries to build the product himself. Both choices tell you the same thing about the old analyst track. The skill that once bought you a seat in investment banking is now valuable because it can teach software how banking works.
Gabe Stengel took a third path. He worked at Lazard after studying computer science at Princeton, then cofounded Rogo in 2022 with John Willett and Tumas Rackaitis. Axios, citing Fortune, reported that Rogo raised $18.5 million in Series A funding led by Khosla Ventures. The company sells generative AI tools for financial services and has described itself as Wall Street's first AI analyst. You can argue with the slogan, and you should, because every startup likes to crown itself early. But the shape of the company is real: a banker and engineers building for bankers, not a general chatbot wearing a finance badge.
The hard part for these startups is not making a demo that looks clever. Banking is full of repetitive work, but it is also full of small judgment calls that are hard to fake. A model can look tidy and still be wrong. It can use the right template and miss the assumption that actually drives the deal. Anyone who has worked around finance knows the spreadsheet is not just math. It is hierarchy, convention, client pressure and fear of sending the wrong file to the wrong person.
That is why former analysts have an opening. They know the boring parts well enough to know which ones can be automated first. Formatting a model, checking links, building a first pass on comparable companies, pulling together a draft analysis, these are obvious targets. Replacing judgment on a live transaction is harder. Frankly, any startup pretending otherwise is selling theater.
Also read: Bhavin Turakhia is spending $30 million of his own money to rebuild Office from scratch • Switch chases a $19 billion valuation as private money floods into AI data centers • Anthropic's Fable 5 Pricing Mess Is Handing Chinese Rivals an Opening
The $22 million round Bloomberg reported is current, and it fits a live market rather than a stale trend. AI companies still need domain experts, and finance remains one of the richest pools of structured, high-value office work. If you are a junior banker, that should make you think carefully about what you are really learning. The hours are brutal. The useful asset may be the map of the work itself.
