Ask Anthropic where enterprise AI stalls and the answer is people, not models. On October 2 the company opened Claude Frontier Academy, a training effort it calls the first of its kind from an AI lab.
The scale behind it is large. Anthropic already runs a Claude Partner Network that it says spans professionals at 46,000 firms holding upward of 175,000 certifications. Those partners have each pledged to certify a set headcount, Accenture to 50,000 staff and PwC to 30,000 among them.
A first intake is already under way, called the Frontier Deployed Engineer Residency. Employers nominate an engineer and name the Claude project that person will lead on return. The format borrows from medical training: instructors from Anthropic run a multi-day session, then participants work a simulated deployment that begins at use-case selection and ends at security review, capped by a graded practical.
Two badges mark progress. The Claude Resident Engineer badge comes first, right after the practical, followed by a 12-week placement in which the engineer runs a real Claude project back at their employer. A second assessment decides who earns the Frontier Deployed Engineer title. The first Frontier Deployed Engineer titles are slated to land in early 2027, according to Anthropic.
Filling out the opening intake are engineers from a roster that runs from Commonwealth Bank of Australia and Novo Nordisk to Accenture, Bain, Capgemini, Deloitte, McKinsey and Morgan Stanley, spanning consulting, banking and life-sciences work. Sessions run in San Francisco, New York and London, and entry is by nomination. The wider ambition is $100M and 10,000 trained engineers by the end of 2027, which Anthropic calls only a first step.
