feat: Add Streamlit dashboard with Blueprint compliance (v2.1.0)

Dashboard Features:
- 8 navigation sections: Overview, Outcomes, Poor CX, FCR, Churn, Agent, Call Explorer, Export
- Beyond Brand Identity styling (colors #6D84E3, Outfit font)
- RCA Sankey diagram (Driver → Outcome → Churn Risk flow)
- Correlation heatmaps (driver co-occurrence, driver-outcome)
- Outcome Deep Dive (root causes, correlation, duration analysis)
- Export functionality (Excel, HTML, JSON)

Blueprint Compliance:
- FCR: 4 categories (Primera Llamada/Rellamada × Sin/Con Riesgo de Fuga)
- Churn: Binary view (Sin Riesgo de Fuga / En Riesgo de Fuga)
- Agent: Talento Para Replicar / Oportunidades de Mejora
- Fixed FCR rate calculation (only FIRST_CALL counts as success)

Technical:
- Streamlit + Plotly for interactive visualizations
- Light theme configuration (.streamlit/config.toml)
- Fixed Plotly colorbar titlefont deprecation

Documentation:
- Updated PROJECT_CONTEXT.md, TODO.md, CHANGELOG.md
- Added 4 new technical decisions (TD-014 to TD-017)
- Created TROUBLESHOOTING.md with 10 common issues

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

config/settings.yaml

@@ -0,0 +1,207 @@
+# ============================================
+# CXInsights - Settings Configuration
+# ============================================
+# Non-secret configuration values
+# Secrets (API keys) go in .env
+# ============================================
+
+# ============================================
+# GENERAL
+# ============================================
+
+project:
+  name: "CXInsights"
+  version: "0.1.0"
+  language: "es"  # Primary language for analysis
+
+# ============================================
+# BATCH PROCESSING
+# ============================================
+
+batch:
+  # Maximum calls per batch (cost protection)
+  max_calls: 5000
+
+  # Maximum audio minutes per batch (cost protection)
+  max_audio_minutes: 40000
+
+  # Default AHT assumption for cost estimation (minutes)
+  default_aht_minutes: 7
+
+# ============================================
+# TRANSCRIPTION (STT)
+# ============================================
+
+transcription:
+  # Default provider
+  provider: "assemblyai"
+
+  # AssemblyAI settings
+  assemblyai:
+    language_code: "es"
+    speaker_labels: true
+    auto_chapters: false
+    entity_detection: false
+
+  # Audio validation
+  audio:
+    supported_formats: ["mp3", "wav", "m4a"]
+    max_duration_seconds: 18000  # 5 hours
+    min_duration_seconds: 30
+
+# ============================================
+# FEATURES (Deterministic Extraction)
+# ============================================
+
+features:
+  # Silence detection
+  silence:
+    threshold_seconds: 5.0
+    min_gap_seconds: 1.0
+
+  # Turn metrics
+  turn_metrics:
+    min_turn_duration_seconds: 0.5
+    interruption_overlap_seconds: 0.3
+
+# ============================================
+# COMPRESSION
+# ============================================
+
+compression:
+  # Target token reduction percentage
+  target_reduction_percent: 60
+
+  # Max tokens after compression
+  max_compressed_tokens: 2000
+
+  # Preserve elements
+  preserve:
+    - customer_intent
+    - agent_offers
+    - objections
+    - resolution_statements
+    - key_timestamps
+
+# ============================================
+# INFERENCE (LLM)
+# ============================================
+
+inference:
+  # Default model
+  model: "gpt-4o-mini"
+
+  # Model settings
+  temperature: 0.1
+  max_tokens: 4000
+
+  # Batch processing
+  batch_size: 10
+  checkpoint_interval: 50
+
+  # Retry settings
+  max_retries: 5
+  backoff_base: 2.0
+  backoff_max: 60.0
+
+  # Response validation
+  require_evidence: true
+  min_evidence_spans: 1
+
+# ============================================
+# VALIDATION (Quality Gate)
+# ============================================
+
+validation:
+  # Confidence thresholds
+  confidence:
+    accept: 0.6
+    review: 0.4
+    reject: 0.3
+
+  # Evidence requirements
+  evidence:
+    required: true
+    min_spans: 1
+    max_span_length_chars: 500
+
+  # Schema validation
+  schema:
+    strict: true
+    version: "1.0.0"
+
+# ============================================
+# AGGREGATION (RCA Building)
+# ============================================
+
+aggregation:
+  # Minimum sample size for statistics
+  min_sample_size: 10
+
+  # Severity score calculation
+  severity:
+    # Weights for severity formula
+    frequency_weight: 0.4
+    impact_weight: 0.4
+    confidence_weight: 0.2
+
+  # RCA Tree building
+  rca_tree:
+    # Minimum percentage to include in tree
+    min_percentage: 1.0
+
+    # Maximum drivers per category
+    max_drivers_per_category: 10
+
+    # Include emergent in separate section
+    separate_emergent: true
+
+# ============================================
+# EXPORTS
+# ============================================
+
+exports:
+  # PDF Report
+  pdf:
+    template: "executive_summary"
+    max_pages: 5
+    include_charts: true
+
+  # Excel Export
+  excel:
+    include_raw_data: true
+    include_pivot_tables: true
+
+  # JSON Export
+  json:
+    pretty_print: true
+    include_metadata: true
+
+# ============================================
+# LOGGING
+# ============================================
+
+logging:
+  # Log level (DEBUG, INFO, WARNING, ERROR)
+  level: "INFO"
+
+  # Log format
+  format: "structured"  # "structured" or "plain"
+
+  # Retention
+  retention_days: 30
+  error_retention_days: 90
+
+  # What to log
+  log_transcripts: false  # Never log full transcripts
+  log_evidence_spans: true
+  log_token_usage: true
+
+# ============================================
+# PROMPT VERSIONS
+# ============================================
+
+prompts:
+  # Active prompt versions
+  call_analysis: "v1.0"
+  rca_synthesis: "v1.0"