Documentation Index
Fetch the complete documentation index at: https://raindrop.ai/docs/llms.txt
Use this file to discover all available pages before exploring further.
Keyword & Semantic Search
Use keyword and semantic search to find specific events and explore data. You can restrict the search to just the user input or the assistant output and apply filters over time, user_id, convo_id etc.
Deep Search
Deep Search is like deep research for your production data. Describe the behavior you’re trying to find and it analyzes millions of interactions to surface matches. It’s the engine that powers custom classifier signals. You start a Deep Search by asking the Triage Agent for a signal in chat (in Slack or the web app). Tell it the behavior you want to track and it’ll scope it, run a Deep Search, seed it with examples it just found, and open the labeling artifact for you to review:Label examples
Mark a few matches as correct or incorrect. Deep Search uses your labels to learn the pattern.

Refine in chat
Tell Triage how to tighten the search (“only count negative mentions”, “exclude URLs with X”) and it’ll adjust the proposal and re-sample. Or hit Update Search in the artifact to regenerate rules from your labels.

Start tracking
Once the matches look right, turn it into a signal. Raindrop trains a small custom model on your labels. It takes about an hour, then runs over every future event and backfills the previous 3 days.

Filtering
Every table in the app has powerful filtering that lets you filter by event, user, conversation, date range, any property, model, tool, and signal.