Predict ride demand and pre-position drivers

Design an AI system for a ride-hailing marketplace that predicts real-time demand and moves driver supply into place before peak congestion hits.

Context

Riders face surge pricing and long waits at peaks; drivers lose money driving empty. Drivers are independent and can ignore suggestions. Supply is limited at peak times.

Clarifying questions to ask first

Which cities and which peaks (commute, events, weather)?
Do we nudge drivers or pay them to move?
What data do we have in real time?

How to approach it

Start from the user problem, not the model: is AI the right tool here?
Describe the experience and where the model fits in it
Define what "good output" means and how you would evaluate it before and after launch
Name the failure modes (wrong answers, bias, privacy, cost) and how you contain them
Close with a launch plan and success metrics

Framework: User and problem -> Is AI the right tool? -> Experience design and human-in-the-loop -> Quality bar and evaluation -> Risks, cost and latency -> Staged launch and metrics

Sample answer

Drivers are the scarce side; a wrong suggestion wastes their time and trust
Optimise for precision over recall when supply is constrained
Start with predictable peaks (airports, events) before general demand

Offline backtest against past peaks
A/B test by city zone: wait time, driver earnings per hour, empty miles
Guardrail: suggestion acceptance rate and driver complaints

Precision-first demand forecasts for predictable peaks, shown to drivers as optional suggestions with expected earnings. Measure rider wait time and driver earnings per hour, guarded by empty miles and suggestion acceptance.

What interviewers look for

Marketplace balance (riders and drivers)
Precision vs recall trade-off
Driver trust and adoption
Guardrails and fallbacks