Oumi's Compounding AI Factory launch highlights a growing enterprise shift: companies want models trained on their own data, not general-purpose AI that hands competitors an edge.
Companies are walking away from off-the-shelf AI models that hand their data to a third party and erase any competitive edge. The alternative – a custom model trained on the company's own workflows – is starting to show real returns.
Oumi, a Seattle startup founded by engineers from Google, Microsoft and Apple, launched its Compounding AI Factory on Tuesday. The platform automates the deployment of a specialized model into production and continuously retrains it on data from real-world use, the company said.
“Every company is becoming an AI company, but nearly all of them are running the exact same closed, generalized models trained on the public web, not on their own workflows, policies, or edge cases,” Oumi CEO Manos Koukoumidis said. Enterprises using Oumi can export the model's weights, its training data and the exact recipe used to build it.
Oumi says a top-five U.S. bank used the platform to modernize 100 million lines of legacy code after a general-purpose AI system failed roughly half its code-translation tests. That figure comes from the company's own case study, not an independent audit.
Morgan Stanley built DevGen.AI in-house
The bank fine-tuned the tool on its own decades-old codebase to translate legacy languages like Cobol into plain English specs developers use to rewrite the code. Since January, the tool has reviewed more than 9 million lines of code and saved the bank's roughly 15,000 developers an estimated 280,000 hours, Mike Pizzi, Morgan Stanley's global head of technology and operations, told The Wall Street Journal.
JPMorgan Chase has taken a similar path at larger scale. The bank has about 450 AI proofs of concept in the works, a number it expects to reach 1,000 next year. Tools like its LLM Suite employee platform and EVEE customer service assistant are already in production, according to Tearsheet.
Mistral, a Paris-based AI company, has built a business around the same premise. Its platform lets companies train systems on their own data. The customer base has passed 100 companies, including HSBC and Stellantis, PYMNTS reported, and revenue exceeds $400 million.
Financial services leads, but data fragmentation slows the shift
Financial firms have moved further on this shift than almost any other sector. They reached high adoption on 27 of 75 AI-supported tasks tracked across eight business functions, according to PYMNTS Intelligence's Enterprise AI Benchmark Report. The report surveyed 60 tech executives at U.S. companies with at least $1 billion in revenue. New AI is broadly deployed or fully embedded in data and technology processes at 81% to 95% of firms surveyed.
Even in that leading sector, 30% of financial services leaders named fragmented or poor-quality data as their single biggest barrier to wider deployment, the same report found. A separate PYMNTS Intelligence survey of executives across industries found 85% describe their data as fragmented or only moderately integrated despite 99% expressing confidence their governance supports enterprise AI.
A company cannot train a specialized system on data it has not organized. That fragmentation makes owning a custom AI model harder than renting one, no matter how fast tools like Oumi's make the technical process.
What the stock data says
Microsoft, whose engineers helped found Oumi, trades at $492.43, down 2.26% on the session. The stock carries an Alpha Score of 72 out of 100, a Moderate rating. JPMorgan Chase sits at $365.18, up 0.87% today, with an Alpha Score of 66, also Moderate.
Both stocks reflect the two sides of the enterprise AI shift: Microsoft supplies the general-purpose models that companies are beginning to question, while JPMorgan is one of the largest customers building its own alternatives. The tension between renting and owning AI will show up in both firms' margins and customer retention metrics over the next several quarters.
Oumi's launch is a reminder that the market for enterprise AI is splitting. The companies that organize their data first will own the sharper models. The rest will keep renting.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.