Project Management View - RogerThattt/Data-Flywheel GitHub Wiki
Phase | Key Activities | Duration
-- | -- | --
Discovery & Planning | Stakeholder alignment, data audit, ROI estimation | 2–4 weeks
Foundation Build | Data lakehouse setup, ingestion pipelines, governance | 4–6 weeks
Model Prototyping | EDA, feature engineering, first ML use case | 3–6 weeks
Operationalization | DLT, CI/CD, model registry, serving infra | 3–5 weeks
Scaling & Iteration | Expand use cases, real-time streaming, user feedback loops | Ongoing
✅ KPIs to Track Churn rate reduction
Mean time to detect/resolution (MTTD/MTTR) on network anomalies
ARPU lift through personalization
Data pipeline latency
Model training & serving throughput
Data quality metrics (completeness, accuracy)