Call QA
PyAI transcription with speaker diarization, then a hybrid deterministic + LLM rubric pass. Every verdict cites a verified line from the real transcript.
Product
Everything CallLoop actually scores, and how
Ticket Audit
Ticket Audit ships with a five-dimension default rubric you can fully customize — Problem Diagnosis, Resolution Correctness, Communication Clarity, Tone & Empathy, Ownership & Handoff Quality — plus a deterministic Response Timeliness metric. Every agent who touched a ticket gets their own independent scorecard, not a shared number. Internal notes are captured and factored in, not discarded. Screenshots are described and placed exactly where they occurred in the thread.
Team Performance
One view across both channels: org and self rollups, a per-dimension pass-rate heatmap, and assignable KPI targets with Met/Missed tracking per agent.
Training
Listen / Drill / Recap — generated from each agent’s own real, stored Top Gap findings. Coaching pulled from what actually happened, not a generic course.
Self-Serve Rubric Builder
Build, version, and activate your own criteria for calls and tickets separately. Save a library, switch what’s active — no engineering ticket required.
Roles
A real Owner / Manager / Agent model. A Manager — not just the original account owner — sees the full team’s results; an Agent sees only their own. Enforced on every route, not a presentation-layer toggle.
How your data stays yours
Two independent layers of tenant isolation — application-layer org identity resolved only from a verified login, and Postgres Row-Level Security re-checking it on every query through a role that cannot bypass RLS even if the app layer had a bug. Every sensitive action — who changed a rubric, who connected an integration, who exported data — is written to a durable, non-editable activity log.
See your own calls and tickets scored — with the receipts.
Request a demo