Laminar

About

Laminar is an open-source observability and analytics platform designed for complex LLM applications and long-running AI agents. It provides deep visibility into agent behavior, enabling developers to monitor, debug, and optimize performance with ease. The platform offers features such as full trace context, allowing teams to understand exactly what an agent was doing at any point and quickly identify where issues occur. With session replay for browser-based agents, Laminar captures screen recordings synchronized with execution traces, making debugging more intuitive and efficient. Designed for fast iteration, Laminar integrates with popular tools like Playwright, Browserbase, and other agent frameworks, helping developers detect patterns, improve reliability, and scale AI systems with confidence. By simplifying tracing and analysis, Laminar reduces the complexity of managing advanced AI workflows.