OpenAI started 2026 the way it ended 2025: buying teams and shutting down their products. On January 8th, the company announced it's bringing on the three co-founders of Convogo, a platform that automated leadership assessments for HR teams and executive coaches. The product itself? Gone.
An OpenAI spokesperson was clear about what they're actually buying. Not the technology. Not the IP. Just the people. Matt Cooper, Evan Cater, and Mike Gillett will work on OpenAI's "AI cloud efforts," whatever that means. The deal was all-stock.
The origin story (actually kind of charming)
Here's the thing: Convogo started because Cooper's mom asked a question. She's an executive coach who hated spending hours writing feedback reports. Could AI handle the boring parts so she could focus on actual coaching?
That was two years ago. Since then, the team built a product that worked with "thousands" of coaches and, per their announcement email, partnered with the "world's top leadership development firms." Client list allegedly includes Amazon, Disney, Adobe. Not bad for a weekend hackathon.
But here's what caught my attention in their goodbye letter. They said the real challenge they discovered wasn't building AI tools. It was bridging the gap between what new models can technically do and what people can actually use them for. That framing probably got OpenAI's attention more than any feature set.
Nine acquisitions in 12 months
This is acquisition number nine since early 2025, according to PitchBook data. The pattern is consistent: bring in the team, kill the product, move on.
Context.ai, Crossing Minds, Roi, Software Applications (the Sky Mac app people): all acquired, all shut down. The talent gets absorbed. The products disappear.
There are exceptions. Statsig, which OpenAI bought for $1.1 billion in September, still serves external customers. And the Jony Ive deal is different altogether: io Products keeps building its own roadmap while collaborating on AI hardware with OpenAI.
But Convogo fits the standard playbook. Small team with domain expertise in making AI actually useful for specific professional workflows. That's the profile OpenAI keeps hunting.
What "AI cloud efforts" might mean
OpenAI won't explain what "AI cloud efforts" actually entails. But the context is interesting.
The company's compute infrastructure runs primarily through Microsoft Azure. They're not building a competing public cloud. So "AI cloud" probably means something like: making enterprise AI deployments work better behind corporate firewalls, handling compliance requirements, that sort of thing.
The Convogo team spent two years figuring out how to turn GPT outputs into usable business processes for non-technical users. That skill set translates.
The real acquisition strategy
OpenAI has now hired teams from: an AI eval platform (Context.ai), a recommendation engine (Crossing Minds), a personal finance app (Roi), a Mac AI interface (Sky), and now an executive coaching tool (Convogo). Plus the big Statsig and io Products deals.
The throughline isn't technical capability. It's applied AI experience. People who've actually shipped products that real users pay for. People who've discovered what breaks when you put a language model in front of someone who doesn't care about AI but does care about getting their job done.
That's harder to find than you'd think. And apparently worth paying in stock to acquire.
The Convogo service shuts down soon. If you're a customer, you'll need to find alternatives and pull your data. OpenAI's M&A team will probably announce another one of these within a few months.
