Skip to main contentSkip to user menuSkip to navigation
System Designer
Back to Practice Hub

Design a Search Ranking System

Build a machine learning-powered search system that delivers highly relevant, personalized results at massive scale with real-time indexing and ranking.

Q: What's the expected scale and types of search queries?
A: Handle 100K+ queries per second across 10B+ documents. Support web search, local search, product search, and entity search with personalization for 1B+ users across 50+ languages.
Analysis: Massive scale requires distributed search infrastructure, multi-tier caching, and efficient ranking pipelines. Different query types need specialized processing: web (semantic + authority), local (geo-relevance), product (conversion signals).
Q: What ranking signals and ML models should be used?
A: Multi-signal ranking: text relevance (40%), authority/quality (25%), user behavior (20%), personalization (10%), business logic (5%). Real-time ML with learning-to-rank models trained on click data.
Analysis: Learning-to-rank requires handling position bias, negative feedback loops, and cold start problems. Need real-time feature engineering, model serving at scale, and continuous online learning from user interactions.
Q: How should the system handle real-time indexing and updates?
A: Real-time indexing with sub-minute freshness for new content. Multi-tier indexing: hot tier for recent updates, cold tier for stable content. Support incremental updates without search disruption.
Analysis: Real-time indexing at scale requires event-driven architecture, write-ahead logs, and careful consistency management. Hot/cold architecture balances freshness with stability while minimizing query latency impact.
Q: What privacy and personalization requirements exist?
A: Privacy-preserving personalization using cohort-based models, differential privacy for analytics, and user control over personalization. Balance relevance with privacy compliance and avoid filter bubbles.
Analysis: Privacy-first personalization: aggregate user behavior into anonymous cohorts, apply differential privacy to training data, implement user controls, and inject diversity to prevent filter bubbles while maintaining relevance.
Q: How should search quality and experimentation be managed?
A: A/B testing framework for ranking algorithms with position bias correction. Relevance evaluation using human judges, click metrics, and satisfaction scores. Guardrail metrics to prevent quality degradation.
Analysis: Search experimentation requires sophisticated statistical methods to handle position bias, network effects, and novelty. Need offline evaluation, online testing, and long-term quality monitoring with multiple metrics.
No quiz questions available
Quiz ID "search-ranking" not found

Interview Practice Questions

Practice these open-ended questions to prepare for system design interviews. Think through each scenario and discuss trade-offs.

1

Global Search Engine Ranking: Design a ranking system for a search engine handling 100B+ documents with sub-200ms query latency globally. Address relevance scoring, personalization, spam detection, and real-time index updates across multiple data centers.

2

E-commerce Product Search: Build a product search ranking system for an e-commerce platform with inventory management, price optimization, seller quality, and conversion optimization. Handle seasonal trends, promotions, and personalized shopping experiences.

3

Enterprise Knowledge Search: Design a search ranking system for enterprise knowledge management with document security, department-specific relevance, expert identification, and collaboration features. Address access control and compliance requirements.

4

Multi-Modal Content Search: Build a ranking system handling text, images, videos, and audio content with unified relevance scoring. Address feature extraction, cross-modal similarity, and performance optimization for diverse content types.

5

Real-Time News Search Ranking: Design a search system for breaking news with recency bias, credibility scoring, fact-checking integration, and trending topic detection. Handle high-velocity content updates and misinformation prevention.

6

Local Business Search Platform: Build a location-based search ranking system with proximity scoring, business quality metrics, review integration, and local intent detection. Address geographic relevance and mobile optimization.