F13 -- Call Intelligence
Analyzes constituent call data to extract trends, common concerns, and sentiment, providing actionable intelligence for representatives and campaign staff.
Key Features
- Transcript analysis: LLM-powered analysis of constituent call transcripts.
- Summary generation: Automatic concise summaries of call content.
- Sentiment classification: Determine caller sentiment (positive, negative, neutral).
- Issue extraction: Identify key issues raised during the call.
- Action item detection: Extract specific follow-up actions from the conversation.
- Satisfaction scoring: Estimate caller satisfaction on a 0.0--1.0 scale.
- Trend aggregation: Track common issues and sentiment trends across many calls.
Server Functions
| Function | Endpoint | Description |
|---|---|---|
analyze_call | call-intel/analyze | Analyze a constituent call transcript |
list_analyses | call-intel/list | List past call analyses |
get_analysis | call-intel/get | Retrieve a specific call analysis |
get_call_trends | call-intel/trends | Aggregate issue and sentiment trends |
Analysis Output
Each call analysis produces:
- Summary: A brief paragraph capturing the key points
- Sentiment: positive, negative, or neutral
- Key issues: List of specific issues raised
- Action items: Tasks that need follow-up
- Caller satisfaction: Estimated satisfaction score (0.0--1.0)
UI Components
- Call intelligence dashboard (
/call-intel): Overview of recent call analyses with sentiment breakdown and trending issues. - Call detail view: Full analysis of a single call with summary, issues, and action items.
- Trend charts: Visual representation of issue frequency and sentiment trends over time.
- Action item tracker: Filterable list of extracted action items across calls.
Database Tables
call_analyses-- transcripts with summaries, sentiment, issues, action items, satisfaction scores