Anthropic said Friday it is putting $100 million into a new training program, the Claude Frontier Academy, with the goal of certifying 10,000 engineers to lead enterprise deployments of its Claude models by the end of 2027. The first track is a residency for what the company calls Frontier Deployed Engineers, and eight large employers have signed on to supply the initial cohorts.

What Happened

The launch partners are Accenture, Bain, Capgemini, Commonwealth Bank of Australia, Deloitte, McKinsey, Morgan Stanley and Novo Nordisk. Each organization nominates its own engineers; there is no open application.

Anthropic describes the format as modeled on medical training. Nominees first attend a multi-day, in-person program with Anthropic engineers in San Francisco, New York or London. That stage ends with a simulated enterprise deployment and a graded practical assessment. Engineers who pass then return to their employers for a 12-week residency in which they lead real Claude projects, followed by a final assessment. Those who clear it receive a Claude Frontier Deployed Engineer badge, which Anthropic expects to begin awarding in early 2027.

RelatedNew York City Comptroller Urges Anthropic to Delay IPO Over Safety and Governance Concerns

The company is looking for hands-on software engineers with strong fundamentals, a record of building with large language models, experience helping colleagues adopt AI and a history of tying technical work to business priorities. Prior experience building AI agents is not required, according to the program page.

Anthropic also gave figures for its existing partner training. It said the Claude Partner Network now spans more than 46,000 firms, whose staff have earned more than 175,000 Claude certifications, and that nearly 4,000 people have completed its Basecamp program.

Why It Matters

The academy is aimed at a specific bottleneck: companies can buy access to frontier models quickly, but turning that access into working systems, especially agents that act inside business software, takes engineers who understand both the model and the customer's processes. The forward-deployed engineer role, popularized by Palantir, places technical staff directly with clients to do that integration work, and AI labs and consultancies have been hiring for it heavily.

By training engineers who already work at consultancies, a bank and a drugmaker, Anthropic is trying to multiply its deployment capacity without hiring all of those people itself. Five of the eight partners are consulting or systems integration firms that sell AI implementation work to other companies, so the credential could also become a selling point in client engagements.

The structure differs from typical vendor certifications, which are usually online exams. Requiring a supervised 12-week residency on a live project means the program will graduate people at a slower pace, but with production experience behind the badge. Reaching 10,000 by the end of 2027 implies cohorts in the thousands within roughly 15 months.

Context

Anthropic has leaned heavily on enterprise customers and partners for growth, and the academy sits on top of the partner network and certifications it already runs. What the residency adds is direct time with Anthropic engineers before nominees lead projects, and a credential tied to an assessed real deployment.

Training of this kind also ties partner engineers to one model family. Graduates will have spent their residency building on Claude, which gives Anthropic an advantage when those engineers recommend tools for their next project.

What to Watch

The first in-person intensives are the next milestone, along with how many engineers each partner nominates. Anthropic has said the residency is the first of several tracks and that it plans to add more roles over time. Whether other employers beyond the eight launch partners gain access, and whether the badge carries weight in hiring outside Anthropic's partner network, will show how far the $100 million stretches.