OpenAI launches DeployCo, a company to embed AI in business workflows

On May 11, 2026, OpenAI announced the launch of DeployCo, a separate company that will help businesses go beyond testing AI and bring it into working systems with measurable results. The aim is to close the most painful stage of the enterprise AI adoption cycle: the move from pilot to routine operation.

Why DeployCo exists

OpenAI says plainly that a strong model alone no longer solves a business's problem. Over the past few years, more than a million organizations have adopted the company's products and API, but the next stage of the market will be decided not by the number of demos and pilots, but by how deeply AI is built into real processes. This is where many companies stall: there is interest and there is budget, but no team that can quickly connect a model to internal data, access rules, controls and employees' daily work.

DeployCo was created to fill that gap. The company will send Forward Deployed Engineers into organizations. These engineers do not work as a remote outside contractor but as part of the client's team. Their job is to work with executives, operations teams and frontline staff to find where AI delivers the most impact, rebuild critical processes around it, and lock in the result in a durable production system rather than another two-week experiment.

How a deployment works

According to OpenAI, a typical project starts with a quick diagnosis: where AI can create the most value, which processes to tackle first and which constraints must not be violated. The team then picks several priority processes, and DeployCo engineers, working inside the company, design, build, test and roll out production systems. They connect OpenAI models to the client's data, tools, controls and business processes so employees can use AI reliably every day.

• Diagnosing the areas where AI delivers the most business value
• Choosing several priority processes together with management
• Integrating models with the client's data, tools and controls
• Testing and launching production systems with clear result metrics

OpenAI also stresses that DeployCo is set up as a standalone business unit but remains an extension of OpenAI itself. That matters for two reasons. First, clients get a single point of contact: they can work with OpenAI, with DeployCo or with both. Second, deployment engineers can build solutions with the future capabilities of frontier models in mind, not just the current product version. For large companies, this is an argument for long-lived systems that won't become obsolete right after the next model release.

Tomoro and the funding

Alongside the DeployCo launch, OpenAI agreed to acquire Tomoro, a consulting and engineering firm that helps enterprise clients turn AI into an operational advantage. The deal gives DeployCo about 150 experienced Forward Deployed Engineers and deployment specialists from day one. It still has to clear standard regulatory approvals and, according to OpenAI's plan, should close in the coming months.

Tomoro already has experience with mission-critical use cases at clients such as Tesco, Virgin Atlantic and Supercell, where what counts is not polished presentations but reliability, integration, risk management and measurable impact in daily operations.

"The task now is to help companies build such systems into the infrastructure and processes that drive the business," said

Denise Dresser, OpenAI's chief revenue officer.

DeployCo starts with more than $4 billion in investment, and OpenAI retains control of the company. The partnership includes 19 global investment firms, consulting companies and systems integrators; OpenAI names TPG, Advent, Bain Capital and Brookfield among the leading participants. Bain & Company, Capgemini and McKinsey are also mentioned separately. The point of this structure is scale: through investment and consulting networks, DeployCo gains access to thousands of companies across industries, while OpenAI gets a channel for turning demand for AI into repeatable deployments, change management and new standards of enterprise automation.

What it means

OpenAI is clearly moving up the stack: it no longer just builds models, it is building a machine for deploying them in large companies. For the market, this signals that the competition is shifting from simple model comparisons to the question of who can package AI into workflows, controls, integrations and KPIs faster. For businesses, the takeaway is even simpler: real value will now come not from pilots but from teams that know how to bring AI to production.