BeyondCX_Insights/docs/TODO.md

# TODO.md

> Lista priorizada de tareas pendientes

---

## Checkpoints Completados

### CP1: Project Setup & Contracts ✅
- [x] Crear estructura de carpetas
- [x] Inicializar repo Git
- [x] Crear requirements.txt
- [x] Crear .env.example
- [x] Crear README.md
- [x] Crear config/rca_taxonomy.yaml
- [x] Crear config/settings.yaml
- [x] Crear schemas Pydantic

### CP2: Transcription Module ✅
- [x] Implementar Transcriber interface
- [x] Implementar AssemblyAITranscriber
- [x] Implementar modelos (Transcript, SpeakerTurn)
- [x] Tests unitarios
- [x] Notebook 01_transcription_validation.ipynb

### CP3: RCA Schemas & Data Contracts ✅
- [x] Implementar CallAnalysis
- [x] Implementar RCALabel, EvidenceSpan
- [x] Implementar Event
- [x] Separar observed vs inferred
- [x] Crear data/examples/

### CP4: Feature & Event Extraction ✅
- [x] Implementar event_detector.py
- [x] Implementar turn_metrics.py
- [x] Tests unitarios

### CP5: Inference Engine ✅
- [x] Crear prompt MAP único
- [x] Implementar LLMClient con JSON strict
- [x] Implementar BatchInference con resume
- [x] Tests de evidence obligatorio
- [x] Notebook 02_inference_validation.ipynb

### CP6: Transcript Compression ✅
- [x] Implementar CompressedTranscript
- [x] Validar reducción >60% tokens
- [x] Integrar en inference
- [x] Notebook 03_compression_validation.ipynb

### CP7: Aggregation & RCA Trees ✅
- [x] Implementar statistics.py
- [x] Definir severity_score con reglas explícitas
- [x] Implementar RCATreeBuilder
- [x] Notebook 04_aggregation_validation.ipynb

### CP8: End-to-End Pipeline ✅
- [x] Implementar CXInsightsPipeline
- [x] Implementar manifests por stage
- [x] Implementar resume
- [x] Implementar exports (JSON, Excel, PDF)
- [x] CLI principal
- [x] Notebook 05_full_pipeline_test.ipynb

### CP-GAPS: v2.0 Blueprint Alignment ✅ (2026-01-19)
- [x] Gap Analysis vs BeyondCX Blueprints (4 docs Word)
- [x] Update rca_taxonomy.yaml with new driver categories
  - [x] churn_risk drivers
  - [x] fcr_failure drivers
  - [x] agent_skills (positive + improvement_needed)
- [x] Update call_analysis.py models with new fields
  - [x] FCRStatus enum
  - [x] ChurnRisk enum
  - [x] AgentClassification enum
  - [x] DriverOrigin enum
  - [x] AgentSkillIndicator model
  - [x] Enhanced RCALabel with origin, corrective_action, replicable_practice
  - [x] Updated CallAnalysis with new fields
- [x] Create prompt v2.0 (config/prompts/call_analysis/v2.0/)
  - [x] system.txt
  - [x] user.txt
  - [x] schema.json
- [x] Update versions.yaml to active v2.0
- [x] Update prompt_manager.py with TaxonomyTexts
- [x] Update analyzer.py to parse new fields
- [x] Update aggregation models and statistics for v2.0
- [x] Update tests for v2.0 compatibility

### CP-DASH: Streamlit Dashboard ✅ (2026-01-19)
- [x] Create dashboard structure (app.py, config.py, data_loader.py, components.py)
- [x] Implement Beyond Brand Identity styling
  - [x] Colors: Black #000000, Blue #6D84E3, Grey #B1B1B0
  - [x] Light theme configuration (.streamlit/config.toml)
  - [x] Custom CSS with Outfit font
- [x] Implement 8 dashboard sections
  - [x] Overview (KPIs, outcomes, drivers, FCR, churn)
  - [x] Outcomes Analysis
  - [x] Poor CX Analysis
  - [x] FCR Analysis
  - [x] Churn Risk Analysis
  - [x] Agent Performance
  - [x] Call Explorer
  - [x] Export Insights
- [x] Advanced visualizations
  - [x] RCA Sankey Diagram (Driver → Outcome → Churn Risk)
  - [x] Correlation Heatmaps (co-occurrence, driver-outcome)
  - [x] Outcome Deep Dive (root causes, correlation, duration)
- [x] Export functionality
  - [x] Excel multi-sheet workbook
  - [x] HTML executive summary report
  - [x] JSON raw data export
- [x] Blueprint terminology compliance
  - [x] FCR: 4 categorías (Primera Llamada/Rellamada × Sin/Con Riesgo)
  - [x] Churn: Sin Riesgo de Fuga / En Riesgo de Fuga
  - [x] Agent: Talento Para Replicar / Oportunidades de Mejora

---

## Alta prioridad (Pendiente)

- [ ] **Run real benchmark with v2.0** - Ejecutar pipeline con 50-100 llamadas reales
- [ ] **Measure actual costs** - Documentar costes reales STT + LLM
- [ ] **Validate v2.0 RCA accuracy** - Manual review de 20 llamadas con nuevos campos
- [x] **Documentation** - Completar stubs en docs/ ✅
- [x] **Test v2.0 with real transcripts** - Validado con batch test-07 (30 llamadas) ✅
- [x] **Update exports for v2.0** - Dashboard incluye todos los campos nuevos ✅
- [x] **Dashboard Streamlit** - Implementado con Beyond branding ✅

---

## Media prioridad (CP9 - Optional)

- [ ] Caching por hash de transcript
- [ ] Batch size benchmarks (encontrar óptimo)
- [ ] Comparar STT providers (Whisper, Google)
- [ ] Comparar LLM providers (Claude vs GPT-4o)
- [ ] DuckDB para analytics de grandes batches

---

## Baja prioridad (Fase 2)

- [x] Dashboard Streamlit ✅ (completado 2026-01-19)
- [ ] Docker containerization
- [ ] CI/CD pipeline
- [ ] API REST (FastAPI)
- [ ] Multi-idioma support
- [ ] Real-time processing
- [ ] Integración BeyondDiagnosticPrototipo
- [ ] Campaign tracking (Blueprint KPI 2)
- [ ] Customer value analysis (Blueprint Pilar 4)
- [ ] Sales cycle optimization analysis

---

## Backlog (Ideas)

- [ ] Automatic prompt tuning based on validation results
- [ ] A/B testing de prompts
- [ ] Confidence calibration
- [ ] Active learning loop
- [ ] Cost anomaly detection

---

**Última actualización**: 2026-01-19 (v2.1 Dashboard + Blueprint Compliance completed)