Sofia Rest builds AI systems that hunt down and fix software bugs. At Sentry, she develops Seer, an AI debugging agent, connecting it to the exact telemetry and context needed to solve issues.
Her work tackles the reality of running AI in production. From telemetry integration to cost tracking, she focuses on practical improvements. She recently shared how her team ran 21 live experiments to improve AI code reviews, proving that simply upgrading to a smarter model isnβt always the right answer.
Before joining Sentry, she studied computer science and cognitive science at Johns Hopkins.