Anthropic platform engineering lead says Claude now writes all the team's code
10/22/2026 — 10/22, 20:19·1 sources (all)·1 reports (all)
Key takeaways
- Katelyn Lesse, who leads Anthropic's platform engineering team, said at Grafana Labs' ObservabilityCON that Claude now generates all the cod…
Background & analysis
On a Tuesday local time, Katelyn Lesse, who leads Anthropic's platform engineering team, said at Grafana Labs' ObservabilityCON event that all of the code her team writes and maintains for its systems is written by Claude. IT之家 reported the remarks on October 22. Lesse oversees Anthropic's platform engineering group, whose scope includes the company's core services, developer tools and product infrastructure. According to Business Insider, the team had roughly 200 people this summer.
Under the arrangement Lesse described, engineers remain responsible for system design and verification, then hand the actual coding work to Claude, with their attention shifting toward system design, validation and code quality. During a large migration of a codebase from Python to Rust, team members built verification mechanisms to confirm the final output met expectations, while AI agents carried out much of the concrete work of developing, monitoring and validating code components.
The team has also begun letting AI agents fix tests on their own, and is exploring ways to allow agents to modify the codebase directly. The public account stops at Lesse's remarks at the event; the team has not disclosed how broadly the practice applies or what comes next.
AI-generated from 1 reports · updated 14 hours ago
Latest turnSpeaking at Grafana Labs' ObservabilityCON, Anthropic platform engineering lead Katelyn Lesse said every line of code her team writes and maintains is now produced by Claude, with engineers shifting to system design, verification and code-quality review. The roughly 200-person team built its own verification checks during a large Python-to-Rust migration while agents handled the bulk of development, monitoring and validation work. The team has also begun letting agents fix tests autonomously and is exploring letting them modify the codebase directly.
Related AI tools
- Heat index
- 65
- All sources
- 1
- All reports
- 1
- First seen
- 15 hours ago
Reports on this story headlines open the original
Speaking at Grafana Labs' ObservabilityCON, Anthropic platform engineering lead Katelyn Lesse said every line of code her team writes and maintains is now produced by Claude, with engineers shifting to system design, verification and code-quality review. The roughly 200-person team built its own verification checks during a large Python-to-Rust migration while agents handled the bulk of development, monitoring and validation work. The team has also begun letting agents fix tests autonomously and is exploring letting them modify the codebase directly.
IT之家AI score 60
Other stories people are talking about
- Heat index 754Anthropic launches Claude Duet agent in research preview14 sources
- Heat index 59Anthropic ships Claude API version 2026-10-21, rejecting several request patterns with 4001 sources
- Heat index 87NewAnthropic to enable auto permission mode by default for Pro, Max and Team1 sources
- Heat index 96NewMIT runs 13 Claude agents unattended on a simulated island1 sources
- Heat index 38Omnara launches open-source alternative to Claude Managed Agents1 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

