Anthropic announced on 2 October a $100 million commitment to train what it calls Frontier Deployed Engineers, with a target of 10,000 of them by the end of 2027. The programme, branded Claude Frontier Academy, will run cohorts in San Francisco, New York and London.
The money is not going on compute or a new model. It is going on people, and that is the interesting part.
How the programme works
There are two stages. The first is a multi-day, in-person course taught by Anthropic engineers, built around a simulated enterprise deployment that runs from choosing a use case through to security review. Engineers who complete it receive a Claude Resident Engineer badge.
The second stage is a 12-week Frontier Deployed Engineer Residency in which participants lead a real Claude project inside their own organisation, supported by Anthropic staff. Completing it earns a Claude Frontier Deployed Engineer badge; Anthropic expects the first residents to finish in early 2027.
This is not an entry-level course. Anthropic says it is looking for hands-on software engineers with strong fundamentals, experience building with large language models, and a record of helping organisations adopt AI.
Who is signed up
The named launch partners are Accenture, Bain, Capgemini, Commonwealth Bank of Australia, Deloitte, McKinsey, Morgan Stanley and Novo Nordisk. The list is mostly consultancies, which fits: these are the firms enterprises already pay to deploy new technology, and Anthropic is effectively paying to make sure the people they send know Claude well.
The argument underneath
Steve Corfield, Anthropic’s global head of business development and partnerships, framed the thesis this way: “A small team of high-agency people with the right skills, access to Claude, and a deep understanding of how their business runs can transform an entire company.”
Read plainly, that is a claim about where the bottleneck sits. Anthropic is not saying its models need to improve before companies see returns. It is saying the shortage is of engineers who can connect a model to a real workflow, with the messy data access, permissions and approval steps that involves. The “forward deployed engineer” model, in which a vendor embeds its own technical staff inside a customer, has become common across AI companies; this programme is an attempt to manufacture those people at a scale no vendor could hire directly.
There is also a competitive logic. Certification creates loyalty: an engineer whose credential says Claude, at a consultancy whose practice is built on Claude, is more likely to recommend Claude on the next project.
What to be cautious about
The supporting numbers in the announcement come from participants and are self-reported. Anthropic has not said how the $100 million breaks down between teaching, residency support and credits, or how many of the 10,000 places go to launch partners. Badges are only as valuable as employers decide they are, and that will take a few cohorts to establish.
Why it matters: if Anthropic is right that people, not models, are the constraint, the next phase of competition between AI labs will be fought over training, partners and certification as much as benchmarks.
Source: Anthropic’s announcement.