Debugging Agents That Think for Hours

Traditional observability tools were built for simple request-response apps. But autonomous AI agents run for hours, generate thousands of execution spans, and make decisions that are nearly impossible to trace after the fact. Laminar is building open-source observability infrastructure purpose-built for this new world — capturing every LLM call, tool use, and function execution with a single line of code.

The $3M seed was led by Atlantic.vc with participation from Y Combinator, AAL.vc, and angel investors including Ben Sigelman (co-creator of OpenTelemetry) and Ant Wilson (CTO of Supabase). Co-founders Robert Kim (ex-Palantir, Bloomberg) and Din Mailibay (ex-AWS) met growing up in Kazakhstan and later studied together at KAIST in South Korea.

See What the Agent Sees

For browser agents, Laminar captures synchronized browser session recordings alongside traces, so developers can see exactly what the agent was looking at when it made a decision. Their Agent Debugger lets developers rerun an agent from any step without starting from scratch, and their Signals feature uses AI to automatically surface failure patterns. Customers already include Browser Use, OpenHands, Rye.com, and Alai.