Design product co-occurrence analytics API
Overview
This question evaluates design skills for RESTful APIs and backend systems, covering API schema and versioning, authentication and rate limiting, caching and cache invalidation, storage models, precomputation pipelines, consistency and staleness bounds, monitoring, and operational concerns.
Snapchat
Sep 6, 2025, 12:00 AM
Software Engineer
Onsite
System Design
0
0
Design RESTful APIs for Co‑Purchase Analytics
Context
You are designing backend APIs to power a "related products" module based on co‑purchase behavior (market-basket analytics). The service must support high read throughput, filtering (time, region, channel), and predictable latency. It will be consumed by internal services and client applications.
Requirements
Design RESTful APIs and the surrounding system to:
-
Endpoints
- (a) Fetch top related products for a given product_id.
- (b) Filter by time window, region, or channel.
- (c) Paginate results.
-
Interface Details
- Specify request/response schemas (including error format).
- Define authentication and authorization.
- Define rate limiting.
- Define caching strategy and cache invalidation.
- Define SLA targets and timeouts.
- Describe idempotency behavior.
- Define versioning strategy.
-
Data & Systems
- Describe the storage model for serving co‑purchase results.
- Describe the precomputation pipeline (batch and/or streaming).
- State consistency expectations and staleness bounds.
- Describe monitoring and alerting.
-
Operations
- Explain how to test the API (functional, performance, data quality).
- Explain how to roll out changes safely (backward compatibility, canary, versioning).
Solution
Login required
Show