Brampton Board of Trade
Agent Performance
Quality and reliability metrics Prudle tracks for every call, day by day.
0
Loading
0
Loading
0
Loading
0
Loading
Per-agent scorecard
Which agent is answering well vs. escalating a lot.
| Agent | Calls | AI Resolved | Routed to Staff | Avg Duration | Avg RAG Lookups | Tool Calls | Tool Errors | Success Rate |
|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||
General
Conversation volume and outcomes, day by day.
Knowledge Base
RAG (knowledge base retrieval) usage and speed.
LLMs
Language model call volume.
Tools
Tool/function call reliability and speed.
Audio & Turn Latency
Per-turn timing metrics (e.g. response time, time to first sentence, turn-taking latency) — names shown are whatever the voice platform reports for your agent, not hardcoded.
No per-turn latency data yet.
Evaluation
Custom evaluation criteria, if configured on your voice agent (Agent settings → Analysis → Evaluation criteria). Shows % of calls passing each criterion, per day.
No evaluation criteria configured on this agent yet.