A forward-deployed engineer, for you

A forward-deployed engineer, or FDE, is a software engineer who works inside your organization instead of at a vendor's office. They sit with the team, learn how the work actually happens, and build AI directly into the workflow. In 2026 it became the most sought-after role in AI. The reason why tells you where the value in all of this really sits, and it is not in the model.

The labs bet billions that deployment is the hard part

Models stopped being the bottleneck. The hard part is wiring one into a specific organization's processes, data, and people, and that takes an engineer on the ground. The major AI labs reached that conclusion at almost the same moment.

Between 11 May and 2 July 2026, four of them stood up deployment organizations worth roughly nine billion dollars combined:

• OpenAI launched the OpenAI Deployment Company on 11 May 2026 with more than $4 billion of initial investment, and agreed to acquire a consulting firm to bring roughly 150 forward-deployed engineers and deployment specialists in from day one. (OpenAI)
• Anthropic launched Ode with Anthropic in May 2026, a $1.5 billion implementation company, as a joint venture with Blackstone, Hellman and Friedman, Goldman Sachs, and others. (TechCrunch, 15 July 2026)
• Amazon Web Services committed $1 billion to a new internal organization for AI-focused forward-deployed engineers, announced 30 June 2026. (TechCrunch, 30 June 2026)
• Microsoft announced Microsoft Frontier Company on 2 July 2026, a new operating business for enterprise AI deployment, with a $2.5 billion investment and 6,000 industry and engineering experts. (TechCrunch, 2 July 2026)

Every company named above is a subject of the reporting, not a client, a partner, or an endorsement. Nothing here implies a relationship with any of them. The four figures were read at source on 18 September 2026: the OpenAI announcement is the company's own, and the Anthropic, Amazon, and Microsoft figures come from trade reporting rather than the companies' own pages.

When every model vendor builds a services army at once, they are telling you something plainly: the value was never in the model. It is in the workflow.

The question to ask: who does your FDE work for?

A vendor's FDE is paid by the vendor, and incentives follow the paycheck. They study your business, and what they learn deepens your commitment to that vendor's platform. The workflow knowledge leaves when they do, and everything they build assumes you keep renting that vendor's models.

None of that is bad faith. It is simply what the job is designed to do. Embedding is not the differentiator. Incentives are.

If the engagement ended tomorrow, who keeps the learning?

Now shrink it to one person

You have the tools. You are not short of access, you are short of someone who will sit down, watch how your Tuesday actually goes, and build the thing into it. What is on offer instead is a course you will not finish, a subscription you already have, or a colleague who is good at this and has eleven minutes.

An independent engineer working for one person looks like this. They are vendor-agnostic, so they pick what fits your problem, including a free tool or no tool at all. They work in your accounts, on your own material, on the real task and not a sample one. And what they build stays with you afterward, because it was always yours.

Nobody is going to deploy a nine-billion-dollar services organization to fix your week. That part scales down to one person, one hour at a time.

The honest limits

This is not a claim that one person with AI equals an engineering team. Some problems need a team, some need software you should simply buy, and some should not be automated at all. A good engineer tells you which of those you have before taking your money for the fourth one.