How joaquinllenado/recurve-ai works
Recurve AI is an autonomous SDR agent: a product description enters through the dashboard, the backend coordinates market research and strategy generation, then validates targeted leads and stores learned corrections. Scout-triggered signals add a separate outreach-pivot path. The dashboard presents strategy, leads, activity, history, and the knowledge graph. Relationships are based on the excerpts, SOURCE INDEX references, and README; unsampled API and UI behavior is represented only where documented, so some links are omitted.
Main entry points: client.ts API client · App.tsx Dashboard · main.py FastAPI routes · strategy.py Strategy generation
Components 27
Dashboard
- Product input
frontend/src/components/ProductInput.tsx - API client
frontend/src/api/client.ts - Dashboard
frontend/src/App.tsx - Strategy panel
frontend/src/components/StrategyPanel.tsx - Lead table
frontend/src/components/LeadsTable.tsx - Activity feed
frontend/src/components/ActivityFeed.tsx - Feed connection
frontend/src/hooks/useActivityFeed.ts - Knowledge graph
frontend/src/components/GraphVisualization.tsx - Strategy timeline
frontend/src/components/StrategyTimeline.tsx - Admin controls
frontend/src/pages/AdminPage.tsx
Backend API
- FastAPI routes
backend/main.py
Agent workflows
- Strategy generation
backend/agent/strategy.py - Lead validation
backend/agent/validation.py
Services and state
- Web research
backend/services/tavily_service.py - SLM inference
backend/services/slm_service.py - Graph access
backend/services/neo4j_service.py - Neo4j store
- Activity broadcast
backend/services/feed_manager.py - Voice transcription
backend/services/modulate_service.py
Scout pivots
- Trigger pivot
backend/agent/scout_trigger.py - Pivot drafting
backend/agent/pivot.py - Signal monitor
backend/workers/scout.py
Other components
- Sales operator
- Tavily
- Fastino Pioneer
- Signal endpoint
- Modulate API
Connections 38
- Sales operator → Product input: provides description
README.md - Product input → API client: submits product
frontend/src/components/ProductInput.tsx - API client → FastAPI routes: calls API
frontend/src/api/client.ts - FastAPI routes → Strategy generation: starts pipeline
backend/main.py - Strategy generation → Web research: researches market
backend/agent/strategy.py - Strategy generation → SLM inference: generates strategy
backend/agent/strategy.py - Strategy generation → Graph access: reads lessons, stores strategy
backend/agent/strategy.py - Web research → Tavily: searches web
backend/services/tavily_service.py - SLM inference → Fastino Pioneer: runs inference
backend/services/slm_service.py - Graph access → Neo4j store: reads and writes
backend/services/neo4j_service.py - FastAPI routes → Lead validation: starts validation
backend/main.py - Lead validation → Web research: fact-checks leads
backend/agent/validation.py - Lead validation → SLM inference: classifies leads
backend/agent/validation.py - Lead validation → Graph access: stores evidence, lessons
backend/agent/validation.py - Strategy generation → Activity broadcast: broadcasts events
backend/agent/strategy.py - Lead validation → Activity broadcast: broadcasts progress
backend/agent/validation.py - Signal monitor → Signal endpoint: polls status
backend/workers/scout.py - Signal monitor → Activity broadcast: broadcasts changes
backend/workers/scout.py - Signal monitor → Trigger pivot: triggers flow
backend/workers/scout.py - Trigger pivot → Pivot drafting: coordinates drafting
- Pivot drafting → SLM inference: drafts outreach
backend/agent/pivot.py - Pivot drafting → Graph access: records pivot
backend/agent/pivot.py - FastAPI routes → Voice transcription: transcribes audio
backend/main.py - Voice transcription → Modulate API: sends audio
- Activity broadcast → Feed connection: streams events
README.md - Feed connection → Activity feed: provides events
frontend/src/App.tsx - Dashboard → Product input: renders
frontend/src/App.tsx - Dashboard → Strategy panel: renders
frontend/src/App.tsx - Dashboard → Lead table: renders
frontend/src/App.tsx - Dashboard → Activity feed: renders
frontend/src/App.tsx - Dashboard → Knowledge graph: renders
frontend/src/App.tsx - Dashboard → Strategy timeline: renders
frontend/src/App.tsx - Strategy panel → API client: fetches strategy
frontend/src/components/StrategyPanel.tsx - Lead table → API client: fetches leads
frontend/src/components/LeadsTable.tsx - Knowledge graph → API client: fetches graph
frontend/src/components/GraphVisualization.tsx - Strategy timeline → API client: fetches history
frontend/src/components/StrategyTimeline.tsx - Admin controls → API client: calls controls
frontend/src/pages/AdminPage.tsx - FastAPI routes → Graph access: serves graph and records
backend/main.py