OpenAI launches DeployCo with $4B to embed engineers inside enterprises

Selling enterprise AI used to be simple: hand over an API key, charge per token, and let the customer figure out the rest. But as enterprise workflows refuse to magically fix themselves, OpenAI is taking matters into its own hands.

OpenAI officially launched the OpenAI Deployment Company, a new standalone business unit dedicated to embedding specialized engineers directly inside major enterprises. Backed by over $4 billion in initial capital, the new entity remains majority-owned and controlled by OpenAI.

The enterprise dream team

To hit the ground running, OpenAI agreed to acquire Tomoro, an applied AI consulting and engineering firm. The deal instantly injects roughly 150 battle-tested Forward Deployed Engineers and specialists into the new company.

Tomoro isn't an academic think tank. Its engineers previously built mission-critical AI systems inside demanding environments for brands like Tesco, Virgin Atlantic, and Supercell.

OpenAI also assembled a formidable coalition of 19 global powerhouses. The roster includes private equity leaders like TPG, Advent, Bain Capital, and Brookfield, alongside tech investors like SoftBank Corp. and Goldman Sachs. Even consulting giants like McKinsey, Bain & Company, and Capgemini are on board.

Together, these partners sponsor or advise thousands of enterprises worldwide, providing an immediate pipeline of corporate workflows waiting to be re-engineered.

Embedding inside the machine

More than one million businesses already use OpenAI products and APIs. However, moving beyond experimental chatbots requires tearing down and rebuilding core operational workflows.

DeployCo's Forward Deployed Engineers will operate inside client organizations rather than offering remote advice. Engagements will start with targeted diagnostics, identifying high-value bottlenecks before embedding engineers to link frontier models directly into proprietary data, internal tools, and strict security controls.

Because DeployCo remains tightly tethered to OpenAI's internal research teams, its engineers can architect systems designed to anticipate upcoming model capabilities, preventing enterprise tech debt before it happens.

Why it matters

We think this is OpenAI's most aggressive commercial land grab to date. Frontier models are rapidly commoditizing at the API layer, and the real enterprise moats belong to whoever controls critical day-to-day operations. By borrowing Palantir's forward-deployed playbook, OpenAI is moving downstream to lock in enterprise recurring revenue.

For enterprise leaders, this solves the painful execution gap between paying for tokens and delivering measurable ROI. Having elite engineers wire proprietary infrastructure directly into next-generation models is compelling. But it comes with obvious risks: handing the keys of your core operational logic to OpenAI creates immense platform lock-in, making it nearly impossible to swap models later.

It also creates friction across the broader ecosystem. While system integrators and consulting firms are listed as partners, OpenAI is now building its own heavyweight professional services arm. Managing high-touch consulting work at scale is notoriously messy, and balancing that against pure software margins will be OpenAI's next major test.