LeadershipNot yet confirmed elsewhere1 publisher2 min readPublished
Anthropic puts $100 million into training other companies' deployment engineers
Anthropic is putting $100 million into training 10,000 engineers from other companies as deployment specialists by the end of 2027. Business Insider ties the program to Claude spending ahead of an IPO, while Anthropic's executives describe a skills gap.
The Board Room · Leadership desk
Three figures: Anthropic hopes to train 10,000 people through the program by the end of 2027; participants then work 12 weeks in residencies at their own companies; $100 million across the 10,000 goal averages $10,000 per engineer, assuming the whole sum goes to trainees.
| Figure | Value | Note | Claim |
|---|---|---|---|
| People Anthropic hopes to train by the end of 2027 | 10,000 people | 3 | |
| Average per engineer, if all goes to trainees | $10,000 per engineer | $100 million spread across 10,000 | 12 |
| Residency at own company | 12 weeks | 2 |
What happened
- Participants attend four-day on-site intensives, then complete 12-week residencies working with Claude tools at their own companies.
- The first cohorts have started in New York and San Francisco, with engineers from Accenture, Bain and Novo Nordisk among them.
- In May, Salesforce, one of Anthropic's major partners, laid out a plan to recruit 1,000 early-career workers for "AI builder" roles.
Why it matters
- decision A company weighing a nomination is choosing to have an engineer spend most of a quarter building with Claude under Anthropic mentors, after Anthropic's own interview picks the person.
- cost At the stated target the program averages $10,000 per engineer on Anthropic's side. What the employer gives up in engineer time was not disclosed.
- contradiction Corfield says the holdup is skills and not motivation, while Business Insider says the aim is higher Claude spending. Neither shows whether the training leads to more Claude spending. Only customer spend data would show that.
- constraint The "gold standard" label arrives before the assessments do, so in our view completion tells an employer little yet beyond the fact that Anthropic accepted the engineer.
The sales reading comes from Business Insider, which wrote that Anthropic is using the program to get more of the corporate world used to AI and to ramp up Claude spending as it approaches its initial public offering [13]. On that reading, a vendor that trains its customers' staff on its own tools is also selling to them. Anthropic's executives describe a problem that exists apart from the sale. "This is a skills gap problem, and that creates a market problem," Steve Corfield, the company's head of global business development and partnerships, said [8]. He told Business Insider that corporate customers wanted a program like this to deploy Claude more effectively, and that motivation is not the holdup [9].
The two accounts fit together. Job postings for forward-deployed engineers have boomed as more companies buy software from AI firms and need help deploying it [11]. A customer that cannot deploy the product cannot expand its use of it, so a vendor that supplies the missing skill also removes a limit on its own revenue.
The money is easier to size than the motive. Spread across the 10,000-engineer goal, $100 million comes to $10,000 per engineer [12]. That is an average, and it assumes the whole sum goes to trainees. The article does not say how the $100 million breaks down, how much of an engineer's time the residency takes, or whether any participating company has committed Claude spending.
The sequencing matters for a buyer. Companies nominate engineers and Anthropic interviews the candidates [5]. Accepted engineers learn on a simulated agentic system they must build and deploy, take a practical test [6], then spend 12 weeks back at their own companies in contact with Anthropic mentors [2]. Twelve weeks is most of a quarter. A company that nominates now gets back, next quarter, an engineer who has spent that stretch working with Claude. We think the likeliest effect is that Claude becomes the default for that engineer's next project. The source does not test that.
The credential is also unfinished. Corfield hopes Anthropic will set "the gold standard for AI skilling" [14], yet Anthropic is still building the program's assessments [7]. Shambhavi Shambhavi, Anthropic's head of go-to-market strategy and operations for partnerships, said they will focus on "what does a great FDE look like" [7]. The first cohorts, in New York and San Francisco, are already running [4].
Anthropic is not the only company moving on this. In May, Salesforce, one of Anthropic's major partners, laid out a plan to recruit 1,000 early-career workers for "AI builder" roles [10]. That design hires its own people, while Anthropic trains other firms' staff.
What to watch
- Whether Anthropic publishes assessment criteria and completion data for the first New York and San Francisco cohorts.
- Whether any Anthropic IPO disclosure breaks out the program's cost or links participating companies' Claude spending to it.
- Whether other model vendors start training customers' engineers, with Salesforce's AI builder hiring as the nearest comparator.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence35
- Adoption12
- Hype gap+30
- Incentives70
- Confidence45
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Anthropic announced a $100 million investment in a forward-deployed engineer training initiative earlier this month.
- [2]
Participants attend four-day on-site intensives and then complete 12-week residencies working with Anthropic's Claude tools at their own companies, staying in touch with Anthropic mentors during the 12 weeks.
- [3]
Anthropic says it hopes to train 10,000 people through the program by the end of 2027.
- [4]
The first cohorts have begun in New York and San Francisco, including engineers from Accenture, Bain, Novo Nordisk and other organizations.
- [5]
Companies nominate engineers for the training, and Anthropic interviews the candidates.
- [6]
Accepted engineers work in a simulated environment, an agentic system they have to learn to build and deploy, and then take a practical test before returning to their companies.
- [7]
Anthropic is still building its assessments throughout the program; Shambhavi said they will focus on "what does a great FDE look like", covering articulation, problem solving and coding skills.
ReportedSupportedSource: Shambhavi Shambhavi, Anthropic head of go-to-market strategy and operations for partnershipsView cited source - [8]
"This is a skills gap problem, and that creates a market problem," Corfield said.
ReportedSupportedSource: Steve Corfield, Anthropic head of global business development and partnershipsView cited source - [9]
Corfield said corporate customers wanted a program like this to help them figure out how to deploy Claude more effectively, and the holdup isn't motivation.
- [10]
Salesforce, a major Anthropic partner, announced a plan in May to recruit 1,000 early-career workers into "AI builder" roles.
- [11]
Job postings for forward-deployed engineers have boomed over the last couple of years, as more companies buy software from AI firms and need help deploying it.
- [12]
$100 million spread across a 10,000-engineer goal averages $10,000 per engineer, assuming the whole sum goes to trainees.
- [13]
Business Insider wrote that, as Anthropic approaches its initial public offering, it is using the new training model to get more of the corporate world used to AI and ramp up Claude spending.
ReportedInsufficientSource: Business Insider's characterization2 sources— create a free account to open themView cited source - [14]
Corfield hopes that, combined with other company initiatives to teach people AI, the program will set "the gold standard for AI skilling."
Sources
1 independent publisher whose own reporting we read for this story.
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
- AI workforce trainingFollow
- Enterprise AI AdoptionFollow
- Forward-deployed engineeringFollow