LangSmith

The product is an observability platform for AI agents and LLM applications that traces execution step by step, monitors performance in production, and helps debug failures while tracking cost and latency. It is for developers, data teams, and operations teams in businesses that build and run agents. It is delivered as a SaaS platform with SDKs for Python, TypeScript, Go, and Java and OpenTelemetry support, with an option to self-host storage inside a VPC.

Key features

• Agent tracing step by step
• Multi-turn message threading
• Real-time production monitoring
• Cost and latency tracking
• Online LLM-as-judge evals
• Code evals and trajectory monitoring
• Webhook and PagerDuty alerts
• Unsupervised topic clustering
• Error analysis templates
• Full-text search on traces
• JSON key-path filtering
• Trajectory queries
• Sub-second query performance
• Self-host inside VPC