Grant access. TrustLens builds the trust layer itself.
No lineage to wire, no answers to push, no tags. The app owner grants read-only access to their data and AI logs; TrustLens’ Discovery Agent reads the retrieval traces and builds the lineage behind every answer. One loader line, and the widget does the rest.
Grant access
App owner adds a TrustLens key and grants read-only access to data sources + RAG/LLM logs — in Studio. No code.
TrustLens deep-dives
The Discovery Agent scans the estate and reads the logs, building the lineage layer for every answer — automatically.
Widget shows trust
One loader line. The widget detects answers on screen and shows the ✓ Verified card. No markup, no per-answer code.
<!-- The only thing the app owner adds. No SDK, no per-answer code. --> <script src="https://app.trustlens.ai/w/YOUR_TENANT.js"></script>
Live — a customer AI app with zero TrustLens code
This chat renders answers as plain text — no data- attributes. The widget (loaded by the one line above) detects each answer and matches it to the lineage TrustLens built from the logs. Ask a question, then click “Why should I trust this?”.