Summary: Peter Bailis, who became Workday’s chief technology officer in May 2025, left last month to join Anthropic as a member of technical staff, focusing on reinforcement learning engineering. This position eliminates a C-suite title for closer technical involvement and places Bailis in a company developing HR software akin to Workday’s offerings.
Anthropic confirmed the hiring, also reported by Business Insider. Bailis’s details have been removed from Workday’s leadership pages and his LinkedIn profile reflects his new role. Previously, Bailis led AI data initiatives at Google Cloud, founded Sisu Data, and was a computer science assistant professor at Stanford. His background in data systems and AI work at Google make him a unique choice for reinforcement learning at Anthropic, emphasizing the company’s focus on bridging model training with product development.
From C-suite to individual contributor
Bailis’s new title, member of technical staff (MTS), is used at both Anthropic and OpenAI, signifying a flat technical hierarchy rather than a step down. This title covers various seniority and compensation levels: MTS base salaries at Anthropic range between $300,000 and $405,000, and at OpenAI from $210,000 to $530,000, often balanced by equity grants.
For Bailis, the shift is clear. A CTO role at Workday, a company with $8 billion in revenue and 18,000 employees, carries authoritative weight, more than an individual contributor at a leading AI lab. However, it offers direct involvement with models, training decisions, and research, without the organizational hierarchy that separates a CTO from the core work. Anthropic, reporting a revenue run rate over $30 billion and engaging over 1,000 enterprise clients each spending more than $1 million annually, has evolved from a research lab to a major player, offering intriguing engineering challenges for ambitious executives.
What makes Bailis valuable to Anthropic now
The Anthropic spokesperson stated that Bailis will focus on reinforcement learning engineering, crucial where reinforcement learning from human feedback is vital for model alignment, and RL-based training techniques enhance agent behaviors and tool usage. Bailis’s expertise with enterprise software—including HR applications that Anthropic is pursuing—not just as an API provider, but as a product company, is crucial now.
The Information revealed Bailis’s appointment aligns with Anthropic’s focus on HR applications. Anthropic already offers Claude plugins for HR tasks like crafting job descriptions and onboarding content. Bailis, experienced in HR and finance software at Google Cloud and Workday, directs this expertise toward advancing Anthropic’s enterprise ambitions.
This points to competitive irony. Workday’s CEO, Carl Eschenbach, mentioned in February 2026 that Workday’s software is internally used by Anthropic, Google, and OpenAI. The company using Workday’s software has now recruited its CTO to develop products rivaling for the same enterprise HR budgets.
Anthropic’s enterprise strategy
Bailis’s recruitment is framed by Anthropic’s rapid expansion in enterprise distribution in recent months. In March, Anthropic introduced a marketplace for Claude-powered enterprise software, letting companies with API commitments buy third-party applications built on Claude, while Anthropic waived the usual revenue share that cloud giants demand. Launch partners were Snowflake, legal AI firm Harvey, and developer