Player Behavior Clustering & Segmentation

You are an analytics engineer building a player behavior clustering and segmentation platform: Game type: [INSERT - mobile, MMO, shooter, subscription] Data scale: [INSERT MAU / events per day] Segment consumers: [INSERT - LiveOps, monetization, design, CRM] Refresh cadence: [INSERT - daily, real-time] Cover the full pipeline: 1. Design - segmentation hypotheses and business questions, actionable vs descriptive segments, persona taxonomy, how teams will consume and act on segments. 2. Feature engineering - engagement, progression, spend, social, skill and content-preference features; time-window aggregates; normalization and seasonality. 3. Modeling - RFM baseline, k-means/lifecycle clustering, sequence and embedding models for behavior journeys, stability selection, interpreters for non-experts. 4. Technology - feature store, batch and streaming jobs, cluster assignment API, orchestration, backfills, data quality monitoring. 5. Data products - segment dashboard, drift and migration tracking (e.g., newbie to whale), overlap analysis, export to push/CRM and experimentation tools. 6. Testing - cluster stability across windows, business-action validation, privacy review, holdouts to prove segment-targeted actions lift KPIs. 7. Launch - training for ops and design teams, naming convention, governance of who can target which segment, rollout by use case.