Head of Product for Claude Code at Anthropic, developing products around changing AI capabilities.[2]
Start with Product management on the AI exponential →
X · _catwu Career
Wu worked as a product engineer at Scale AI in 2018–2020 and an engineering manager at Dagster Labs in 2020–2021, where she led its first commercial product. She was a partner at Index Ventures from October 2021 through August 2024. She joined Anthropic that August on its Research PM team, then led product for Claude Code and Cowork.[3]
Ideas
Wu’s product-management essay follows one repeated task: asking successive models to add a table tool to Excalidraw. Earlier models failed; later ones made the task practical to demonstrate. She uses that changing baseline to explain why teams should prototype quickly, establish whether an agent can perform a task before optimizing its cost, and remove workarounds when a stronger model makes them unnecessary. Her examples include using Claude Code for user-feedback analysis and evaluation tools.[2]
Tools, writing & talks
Writing, research & talks
Start here
Product management on the AI exponentialclaude.comWu’s first-person account of prototyping, repeated capability tests and product decisions at Anthropic.
Start with Wu’s essay to see how repeated software tasks can test a model’s changing capability. It assumes familiarity with product development but requires no installation.
Author[2]a retrospective on making claude code from our early team! https://t.co/Kk9mz4BYXx
Catalog workOriginal source- Claude Code Liveresources.anthropic.com
An Anthropic webinar with Wu, Boris Cherny and Cal Rueb introducing Claude Code and answering questions about the coding workflow.
Co-presenter[4]