Anthropic's engineering and product heads explain how Claude is used to build Claude
Overview
Latent Space interviews Anthropic's heads of engineering and product, Katelyn Lesse and Angela Jiang, on how Claude Opus 5.5 was launched and built.
The interview covers Managed Agents, which keep the session log and harness separate from the sandbox that runs code, so a failed sandbox does not end the agent session and credentials can stay out of generated code's reach.
A Latent Space post says Claude is used to build Anthropic's first-party tools Code, Tag, and Duet, and that more than 50% of Anthropic's top customers use the platform. The post is promotional, and the report does not verify these figures.
Written by AI from the articles below · updated Oct 27, 3:41 PM ET
Check the sources:
Developments
2 developments
- Oct 27, 2:58 PM ET · 1 articleAnthropic's engineering and product heads explain how Claude is used to ship ClaudeLatent.Space: Anthropic's engineering and product heads explain how Claude is used to ship Claude
- Oct 27, 2:40 PM ET · 2 articlesAnthropic engineers explain how the Claude Platform ships Opus 5.5 safelyLatent Space: Anthropic separates Claude's brain from its hands in agent architecture
Article timeline
The articles in this story. Times are ET.
Latent Space: AI News & Views@latentspacepodXAnthropic's engineering and product heads explain how Claude is used to ship ClaudeAILatent Space announces a behind-the-scenes interview with Anthropic's heads of engineering and product, Katelyn Lesse and Ang Jiang, about the launch of Opus 5.5. The post says the platform is used to build first-party tools such as Code, Tag, and Duet, and is used by more than 50% of Anthropic's top customers.
- Latent SpaceBlogPickAnthropic separates Claude's brain from its hands in agent architecture
AILatent Space interviews Anthropic's heads of Engineering and Product about how Claude Opus 5.5 was launched and built. The discussion covers Managed Agents, which keep the session log and harness apart from the sandbox that executes code, so a failed sandbox does not end the agent session and credentials can stay out of generated code's reach.
- Latent SpaceBlogAnthropic engineers explain how the Claude Platform ships Opus 5.5 safely
AIKatelyn Lesse and Angela Jiang explain the engineering behind the Claude Platform in this Latent Space interview. Topics include launching Opus 5.5, separating the model's brain from the agent's hands, scaling APIs, and building managed agents. The episode also covers Anthropic's S-1 and the platform's role in its business.
Heat Trend
- Current comparable heat
- 14
- Peak comparable heat
- 19Oct 27, 4:00 PM ET
- Change in past 24 hours
- –
The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.