Design an API for pay computation with retries
Company: DoorDash
Role: Software Engineer
Category: Software Engineering Fundamentals
Difficulty: medium
Interview Round: Online Assessment
You’ve implemented logic that computes a dasher’s pay from order event logs (events include timestamps and state transitions).
Design an HTTP API to expose this as a service. Cover:
- Endpoints (e.g., compute pay for a dasher over a time window; submit events; query results)
- Request/response schema (including datetime formats)
- Idempotency strategy (to safely handle client retries)
- Failure handling and retry semantics (client vs server retries, timeouts, backoff)
- Validation and error codes
- Observability (logging/metrics/tracing)
Assume this API will be called by other internal services and must be robust to duplicate requests and partial failures.
Overview: Assesses the candidate's ability to design robust HTTP APIs covering endpoint modeling, request/response schemas, datetime formats, idempotency strategies, retry semantics, validation/error codes, and observability for pay-computation services.
Read the full DoorDash Software Engineer interview experience this question came from
Community answers
Answer by gabrielalveskuabara
Requirements:
Payout amount for orders
Calculate payout for each event (Define rules)
Only finished events should be considered
Parameters:
driver_id
time_window
Response:
HTTP response with float amount
API request for dependecies:
Has driver_id and time_windows to query all orders started and finished in this window
Idempotency:
After calculating payout, mark orders as paid
Failure handling:
Set retries, timeouts and backoff in case request fails