Overview
Fleet health for quality, speed, and cost across your agents.
The Overview page (/overview) is the home view for organization-wide health after you sign in.
What you see
Overview groups metrics into three lenses:
- Quality — Outcomes such as resolved, abandoned, and escalated rates, plus goal and evaluation signals when available. Resolved aligns with completed sessions.
- Speed — End-to-end reply latency (p50 / p95), Over SLO share, and reply latency trends for the selected window.
- Cost — Estimated large language model (LLM), speech-to-text (STT), and text-to-speech (TTS) spend, including cost per successful outcome when data is present.
Charts and insight cards update with the filters you choose. Use Overview to spot regressions, then open Agents or Sessions for the underlying traffic.
Filters
Narrow the fleet view by:
- Date range
- Agent (deployment)
- Agent version
Filter choices apply to the cards and charts on the page. Clear or change them when you want a wider picture.
Where to go next
- Open Agents when a single deployment looks unhealthy.
- Open Sessions when you need a specific call or chat.
- Open Topics when you care about volume and outcomes by subject.
- Open Cost & margin when spend looks wrong or models need rates.