Inspect source references, policy events, model usage and outcomes in per-answer Decision Records. Export evidence under access and retention controls.
Review the topic queried, cited documents, transcript references, structured records and computed evidence.
Inspect prompt versions, privacy checks, connector approvals, routing outcomes and recorded fallbacks.
Export supported JSON or PDF records with integrity checks and access controls. Retention and signed-link settings depend on configuration.
Assemble reviewable evidence and governance records across native chat, the web widget, Slack, Microsoft Teams, Chatwoot and API conversations. Inspect how an answer was produced without exposing a model's hidden reasoning.
Export Decision Records as JSON or PDF packages with integrity checks, controlled access and configurable retention. Conversation evidence bundles support signed, expiring links and SIEM-ready audit records.
Run evidence checks in shadow mode or enable filtering of confidently irrelevant material while retaining protected evidence. Low sufficiency is not a new automatic refusal gate. Optional native-chat quality scores and post-answer grounding audits inform administrators without rewriting the delivered answer.
Use versioned prompts, access boundaries, moderation, injection controls, output scrubbing and permission-gated grounded inference. Govern external tools and destinations. Append-only, hash-chained events retain configuration changes, sources, decisions and usage telemetry.
Architecture and controls depend on the deployment. Review source quality, permissions, privacy, model failure, retention and export behaviour against your own requirements.
Bring representative sources, a useful business question and the controls your organisation needs.
