How Cleric’s AI SRE leveled up with continuous learning through LangSmith

Cleric is an AI agent designed to assist engineering teams in debugging production issues by conducting parallel investigations using existing observability tools and infrastructure. It communicates with teams through Slack, sharing findings and asking for guidance when needed. LangSmith helps monitor and compare the performance of different investigation strategies running simultaneously, providing clear visibility into these parallel investigations and experiments. Cleric learns continuously from interactions within each customer environment and uses LangSmith to manage this continuous learning process by creating generalized memories that strip away environment-specific details while preserving core problem-solving approaches. This system allows Cleric to maintain separate knowledge spaces for unique environments and procedures, alongside a growing library of generalized problem-solving patterns that benefit all users.