▶ Watch ↗18:57
Conference affiliation: Director of Field Engineering · Sourcegraph · 2026
Andrew Norrish is director of field engineering at Sourcegraph, working with enterprise engineering teams on understanding large codebases. His focus is the investigation that follows code generation: determining what a patch affects, tracing a production failure to the deployed code, and finding where a repair must reach across repositories and services.
Before Sourcegraph, Andrew led healthcare software projects, delivering large deployments to hospital organizations across North America. His earlier work at Sourcegraph included technical project management. Based in Denver, he returned to Colorado after several years in the Midwest.
In his talk on AI-generated code and review capacity, Andrew describes a shift in customer conversations: questions about reducing headcount gave way to concerns about reviewing the growing volume of code and investigating failures after deployment. A reviewer assessing a patch and an SRE responding to an incident both need context beyond the file in front of them—callers, dependencies, and changes elsewhere in the system. He explains how code search and a code graph can help them follow those connections.
Two customer examples clarify the decisions he emphasizes:
▶ Watch ↗18:57
Loading more talks and highlights…