Judging Criteria and Winner Selection:

• Live question accuracy (30%) - judges ask your Genie space five questions you haven't seen
• Data quality (25%) - clean joins, documented tables, judgment calls written down
• Interactive dashboard (20%) - a stranger types their ZIP and gets something true, including for ZIPs with almost no data
• Defensible metric (15%) - normalized rather than raw counts, and you can explain what it misses
• Story (10%) - four minutes, clear finding, no jargon

Prizes

• Grand Prize: Lululemon Backpack + Databricks Water Bottle (Top 3 Teams)

Rules

Teams will make different judgment calls about what counts as a rodent complaint and which ZIPs to exclude. This is expected and allowed. There is no single correct answer. You are judged on whether your choices are defensible and documented, not on matching a reference solution.