Todd Underwood
Todd Underwood is a reliability engineering leader who served as head of reliability at Anthropic, where he built the AI Reliability Engineering (AIRE) team focused on the availability, latency, and quality of Claude models and infrastructure. He previously spent nearly 15 years at Google, where he founded the Machine Learning Site Reliability Engineering (ML SRE) organization, and briefly led reliability for the Research Platform at OpenAI. He is a co-author of the O'Reilly book Reliable Machine Learning: Applying SRE Principles to ML in Production (2022). He announced his departure from Anthropic in early 2025 and is currently taking time off.
“You can imagine that people interacting with an agent, and you get access to someone's source code, you can potentially get secrets.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.