Demand forecasting and AI-powered rebalancing recommendations
What feeds the AI Ops demand forecast — historical rides, weather from Tomorrow.io, events from PredictHQ, and the holiday calendar — and what happens when a source goes down.
How to read the H3-hex demand forecast on the Heat Maps page — layers, horizons, the time slider, and tile tooltips.
Common questions about Levy AI Ops — accuracy, privacy, pricing, customization, and what it does and doesn't do.
Per-subaccount controls for AI Ops — ai_ops_enabled, ai_ops_tier, tech_cost_per_mile_usd, rebalance_battery_floor, and what each tier unlocks.
Enable AI Ops for a subaccount, run the first backfill, and verify the forecast surface is live.
The technician's view — how the Route tab in the operator-app shows today's stops, handles offline completions, and supports abandoning a route.
Demand forecasting, rebalancing recommendations, and joint swap+rebalance routes — the AI layer on top of Levy Fleets.
Read the ranked card grid on /dashboard/operations/rebalance — how lift is calculated, when to Accept, Snooze, or Dismiss, and how acknowledgements feed the model.
How AI Ops packages low-battery swaps and rebalance moves into a single VRP-optimized route for each technician — including auto-maintenance and abandon flows.
Common problems with the AI Ops demand surface, recommender, and technician routes — and how to diagnose them.
What an unmet-demand event is, how Levy AI Ops captures it from rider app sessions, and how to use the unmet-demand heatmap.