I build AI agents for clients. The question that kept coming back was always the same: "what did the bot actually say?" I had traces. My clients had a black box. So once a month I hand-assembled transcripts and screenshots into a report nobody fully trusted, because by the time they read it, it was already stale. Conversation Explorer is the thing I got tired of rebuilding. A live conversation viewer you can embed. React component, one script tag, or an iframe. Your client sees the conversations happening now, not an export from last Tuesday. On the Agency tier it's white-label, so it looks like your product, not like a tool you're renting. Reward signals from those conversations. Voice-of-Customer themes with sentiment, an Agent Quality score from 0 to 100 with friction attributed three ways (user, agent, system), and a Feedback Batch pass that clusters thumbs-down conversations into failure modes and suggests a prompt diff. That diff is a suggestion. A human reads it and applies it. The product never trains models and never edits your prompts on its own. Measurement loop, human in control. Ingestion is plain REST. It doesn't care what framework you're on. Human to human, agent to human, agent to agent, all the same shape. Two things I'd genuinely like feedback on: 1. The embed. Wire the viewer into your own agent and tell me where it breaks. Theming and origin allowlisting especially. 2. The quality score. Run it on conversations you already know are good and bad. Does the 0 to 100 land where your gut says it should? Does the friction attribution point at the right culprit? That's the part I most want to be told I got wrong. Free tier is 25,000 credits a month, renewing, with unlimited workspaces and conversations and 30-day retention. The free tier is the offer. I'll be around all day. Ask me anything, including what it doesn't do yet.

Forum Threads

p/conversation-explorer

•

5d ago

Is your agent working at its best?

If you live with that question then you ll love Conversation Explorer.

We want to help builders to get their full potential from their AI Agents.
We ll support you while iterating and provide a score you can optimize and measure quality over time.

View all

Conversation Explorer

AI Agents Observability & Reinforcement Learning

AI Agents Observability & Reinforcement Learning

Forum Threads

Is your agent working at its best?

Forum Threads

Is your agent working at its best?