ABAB Pedia

Boris Cherny

Boris Cherny is a software engineer at AnthropicAnthropicAnthropic is an artificial intelligence research and product company founded in 2021 that develops the Claude models, assistant, and tools such as Claude Code, and operates as a public benefit corporation. Dario Amodei and Daniela Amodei are its co-founders, and its research covers model safety, interpretability, and steerability.Open the full entry who created and leads the coding agent tool Claude Code. Before Anthropic he worked at MetaMetaMeta (Meta Platforms, Inc.) is a publicly listed American technology company, renamed from Facebook in 2021, that operates social communication, digital advertising, virtual reality, smart glasses and artificial intelligence businesses.Open the full entry, where he was tech lead for Facebook Groups and later worked on Instagram, and he is the author of the TypeScript book Programming TypeScript.

Contents12 sections
Key facts

Early Career: Venture Capital, Ad Tech and Startups

Early in his career Cherny worked in venture capital and ad tech and at several startups, most of which no longer exist.⁠[1]

Before joining Meta he also worked in finance.⁠[2]

Programming TypeScript

In May 2019 O’Reilly Media published Cherny’s book Programming TypeScript, a 322-page title aimed at intermediate to advanced readers.⁠[3]

The book’s author note described him at the time as an engineering and product leader at Facebook, interested in programming languages, code synthesis and static analysis, who ran the San Francisco TypeScript Meetup in his free time.⁠[1]

Working for Instagram from Japan

At Meta, Cherny served as tech lead for Facebook Groups, commuting to the company’s Menlo Park offices in California.⁠[4]

In December 2023 he wrote on his blog that for the previous year and a half he had been working for Instagram from Japan, in a time zone shared by few other Meta engineers; during this period his focus shifted toward simplifying the codebase and improving infrastructure, and he mentored engineers in the Singapore office.⁠[4]

Joining Anthropic and Prototyping Claude Code

Cherny joined Anthropic in 2024, working on a team at the intersection of product and model development.⁠[5]

Within about a month of experimenting, he built a prototype: a simple terminal tool with no graphical interface that used the Claude API to help write and modify code, which he named Claude Code. He shared it with his team the next day, colleagues began using it for daily work, and the tool then spread organically across teams inside the company.⁠[5]

Public Release of Claude Code

On February 24, 2025, Anthropic introduced Claude Code alongside Claude 3.7 Sonnet as a command line tool for agentic coding, offered as a limited research preview that could search and read code, edit files, and write and run tests.⁠[6]

On May 22, 2025, with the launch of Claude 4, Anthropic made Claude Code generally available and released beta extensions for VS Code and JetBrains as well as the Claude Code SDK.⁠[7]

Lenny’s Podcast Interview

On February 19, 2026, Lenny Rachitsky’s podcast published an interview with Cherny, introduced as “Head of Claude Code.” The episode covered how Claude Code grew from a quick hack to 4% of public GitHub commits, why he believes coding is “solved,” and why he briefly left Anthropic for Cursor before returning after two weeks.⁠[8]

The Pragmatic Engineer Podcast

On March 4, 2026, The Pragmatic Engineer podcast hosted by Gergely Orosz released a conversation with Cherny. The episode notes describe him as having spent five years at Meta as a Principal Engineer, where he led causal analysis of how code quality affects engineering productivity.⁠[9]

At Meta he moved to Instagram and later to developer infrastructure, leading migrations from Python to the Facebook monolith and from REST to GraphQL; he was also one of Meta’s most prolific code reviewers and wrote lint rules to automate recurring review comments.⁠[9]

The episode also traced how Claude Code grew from a side project into a core internal tool at Anthropic. On his working method, he said: “once there is a good plan, it will one-shot the implementation almost every time.”⁠[9]

Talk at STATION F

On March 11, 2026, Cherny spoke at STATION F. He said Claude Code was released publicly in the February after the prototype, and that each new model improved the tool and drove further adoption, adding: “For us, the model is the product and the product is the model.”⁠[5]

He recalled that he initially thought top-down enterprise sales would be the right strategy, but Claude Code instead spread into companies through individual developers trying it first.⁠[5]

He said he had not written code by hand since the previous November, instead running multiple parallel Claude Code sessions to generate, test and modify software. He said coding was essentially solved for codebases resembling the training data, such as React and TypeScript, and predicted that software engineering would be largely automated by the end of the year.⁠[5]

He said his biggest concern was safety, and that if capability outpaced safety, Anthropic would rather slow product development.⁠[5]

May–June 2026: Code with Claude and Fortune Brainstorm Tech

On May 19, 2026, Cherny spoke at the Code with Claude conference in London.⁠[10]

In June 2026, at the 25th annual Fortune Brainstorm Tech conference in Aspen, he said that that morning he was “managing maybe a few hundred” agents, whereas about a year and a half earlier developers used “one tab of Claude Code in one window”; Claude Code now uses subagents, with another Claude instance rather than a human writing the prompts.⁠[10]

He compared the leap in coding productivity to the invention of the printing press and said that “the role of a builder is just totally changing.”⁠[10]

Blog Post and Computer History Museum Conversation

On September 19, 2026, Cherny published “I am often wrong” on his blog, describing a six-step problem-solving approach he had shared with his team: understand the available information, gather missing information, define the problem, define a clear and simple approach, define a goal, and act urgently; he wrote that new information sends him back to revise the problem, approach and goal, adding: “When there's new data, you have to update your priors.”⁠[11]

On September 23, 2026, the Computer History Museum in Mountain View, California, hosted “In Conversation with Boris Cherny, Creator of Claude Code,” listing him as Creator and Head of Claude Code at Anthropic.⁠[12]

Related Organizations, Websites, and Public Accounts

Sources

  1. About the Author (opens in a new window)
  2. Boris Cherny (opens in a new window)
  3. Programming TypeScript (opens in a new window)
  4. Learning to work (very) remotely (opens in a new window)
  5. Boris Cherny, Anthropic: “I have not written a single line of code since November” (opens in a new window)
  6. Claude 3.7 Sonnet and Claude Code (opens in a new window)
  7. Introducing Claude 4 (opens in a new window)
  8. Head of Claude Code: What happens after coding is solved | Boris Cherny (opens in a new window)
  9. Building Claude Code with Boris Cherny (opens in a new window)
  10. The Man Behind Anthropic’s Claude Code Hasn’t Written a Line of Code in 8 Months — Here’s What He Does Instead (opens in a new window)
  11. I am often wrong (opens in a new window)
  12. In Conversation with Boris Cherny, Creator of Claude Code (opens in a new window)
  13. Boris Cherny's Blog (opens in a new window)
  14. About (opens in a new window)