Startup · Bengaluru
Ola interviews cover ride-hailing domain challenges, backend systems at scale, and data analytics for mobility products.
Process: Online Assessment → Technical × 2 → HR
What engineering challenges are unique to a ride-hailing platform like Ola?
Tip: Ola: real-time matching of 1M+ drivers and riders, dynamic pricing (surge), GPS-based ETA calculation, routing, payment, driver earnings management. Engineering: location data at massive scale (100K+ GPS updates/sec), global dispatch optimisation (NP-hard problem solved heuristically), and two-sided marketplace dynamics.
Design Ola's real-time driver-rider matching system.
Tip: Driver state: location (updated every 3s), availability, vehicle type, rating. Matching: geospatial index (H3 hex grid or geohash) to find nearby drivers → score by ETA, rating, acceptance rate → offer to top N drivers with 15-second timeout. Fallback: expand search radius. Async: no blocking — offer/accept via separate channels.
How does Ola calculate surge pricing and what engineering system enables it?
Tip: Surge: demand / supply ratio per geo-zone (1km² hexagon). Compute every 1 minute: open requests in zone / available drivers in zone. Multiplier tiers from 1.0× to 3.0× based on thresholds. Infrastructure: ride request stream → Kafka → Flink aggregation per zone → Redis pub/sub to pricing service → update driver app.
Write code to find the shortest path in a weighted graph using Dijkstra's algorithm.
Tip: Dijkstra: min-heap priority queue. Start from source (distance 0), relax edges, push (new_dist, neighbor) to heap only if improvement. O((V+E) log V). For Ola road network: road segments are edges with time-weighted costs. Handle negative weights: Bellman-Ford (not Dijkstra). Always clarify this.
How would you store and query real-time driver locations efficiently?
Tip: Redis GEOADD/GEORADIUS: store (driver_id, lat, long), query radius in O(N+log M). Alternative: H3 hexagonal grid — bin drivers into cells, query cell + neighbours. At 1M drivers × updates every 3s = 333K writes/sec — Redis cluster is essential. Drivers update only when location changes significantly (>50m).
What is the difference between REST and gRPC? When might Ola use gRPC internally?
Tip: REST: text-based (JSON over HTTP/1.1), human-readable, wide tooling. gRPC: binary (Protocol Buffers over HTTP/2), strongly-typed contracts, bidirectional streaming, 5–10× lower latency. Ola internal services (matching engine ↔ pricing service ↔ driver state) use gRPC for low-latency inter-service calls. REST for external-facing APIs.
Tell me about a time you designed a system for a scenario you had never encountered before.
Tip: Ola operates at a scale few engineers have personal experience with. Show your first-principles reasoning: how you broke the problem into known sub-problems, what assumptions you validated, and how you iterated on your design. Intellectual curiosity and structured reasoning matter more than perfect answers.
How would you implement a cancellation penalty system for riders who cancel trips frequently?
Tip: Track: cancellation rate per user over rolling 30 days. Tier system: >20% cancellation → warning notification; >30% → ₹25 penalty charged. Grace: first 2 cancellations/month free. Technical: event-driven (TRIP_CANCELLED event) → increment counter in Redis → check threshold → trigger penalty via billing service.
What is geofencing and how does Ola use it operationally?
Tip: Geofencing: define virtual geographic boundaries (polygons) and trigger actions when an entity enters/exits. Ola uses it for: airport zones (no pickup without permit), restricted zones (no drop at certain areas), demand zones (surge activation boundaries), driver home zone definitions. Point-in-polygon check: ray casting algorithm.
Explain how Ola's driver earnings and incentive settlement works.
Tip: Earnings: base fare + per-km + per-min + surge multiplier − Ola commission (20–25%). Incentives: daily guarantees, peak-hour bonuses, streak bonuses. Settlement: daily automated via NEFT/UPI to driver bank account (T+1). Technical: ride completion event → earnings calculator → wallet credit → incentive eligibility check → settlement batch.
ANI Technologies (Ola) interviews follow a 3-round process. Here is what to expect and how to prepare for each stage.
Upload your resume and get questions scored across technical depth, communication, structure, confidence, and relevance — the same criteria Ola panels use.