Agent observability and evals: 8 companies ranked

Companies selling tracing, monitoring and evaluation of AI agents in production, in SalesBook agentic-sector rank order. Every company has a profile with dated public sources and a live check of its agent interfaces (MCP, llms.txt). Ranks use the published SalesBook method; tell us who is missing.

• Braintrust#7Observability and evaluation platform for AI agents in production: tracing, LLM/code/human scoring, experiments and datasets for engineering and product teams.
  MCP connector: nollms.txt: no
• Comet (Opik)#10Opik, an open-source LLM and agent observability and evaluation platform, plus ML experiment management, sold to AI/ML teams and enterprises.
  MCP connector: nollms.txt: yes
• Raindrop#18Monitoring and issue detection for AI agents in production: tracing, silent-failure detection, triage and simulations for teams shipping agents.
  MCP connector: yesllms.txt: yes
• Confident AI#30LLM evaluation, observability, red teaming and AI governance platform for enterprise teams; maker of open-source DeepEval and DeepTeam.
  MCP connector: yesllms.txt: yes
• Laminar#40Open-source (Apache 2.0) observability platform for AI agents: tracing, replay/debugging, failure-pattern signals and evals; hosted or self-hosted.
  MCP connector: nollms.txt: no
• HoneyHive#75Observability and evaluation platform for AI agents: tracing, online/offline evals, prompt management; SaaS, hybrid or self-hosted for enterprises.
  MCP connector: nollms.txt: yes
• Prefactor#99Observability, evaluation and identity controls for production AI agents.
  MCP connector: nollms.txt: yes
• AgentOps#113Observability and debugging platform for AI agents, used with frameworks such as CrewAI.
  MCP connector: nollms.txt: no