#11Typeahead Search Autocomplete
11 / 20
WorkflowCanvas
Google logoGoogle System
MediumFundamentals
30–45 min

Typeahead Search Autocomplete

Design a high-availability typeahead suggestion service returning top 5 search completions within 20ms of every keystroke using distributed trie structures.

Scale & Traffic5 Billion queries/day · 20ms keystroke SLA
Monthly Cloud Budget$10,000/month
Latency Target (P99)< 50ms Realtime SLA
Availability SLA99.99% Multi-AZ Active
OPEN IN PLAYGROUND CANVASVIEW REFERENCE WORKFLOW

System Requirements

  • Real-time suggestion response in < 20ms
  • Prefix matching ranked by global & personal query popularity
  • Offline frequency aggregation and daily dictionary rebuilds
  • Edge caching of top 100k hot prefixes

Graded Constraints

01Cost limit: $10,000/month
02CDN edge caching for popular single & double letter prefixes
03Dedicated in-memory Trie index servers
04Batch stream processing to update query frequency tables

Ready to design?

Open this challenge in the interactive system design playground and satisfy all architecture constraints.

OPEN IN PLAYGROUNDVIEW WORKFLOW
Explore More

Related Architecture Challenges

All Problems
BitlyEasy

URL Shortener

Build a high-performance URL shortening service like TinyURL or Bitly that generates 7-character aliases, handles 100:1 read-to-write ratios, and caches hot redirect lookups.

100M URLs/month →
UberMedium

Scalable Notification System

Architect a multi-channel notification engine supporting iOS APNS, Android FCM, SMS (Twilio), and transactional emails with priority routing and deduplication.

10M+ notifications/day →
AmazonMedium

Distributed Cache Cluster

Design a distributed in-memory key-value caching system supporting consistent hashing, replication, write-through/write-back policies, and LRU eviction.

1M operations/sec →
Typeahead Search Autocomplete — System Design Problem | Mimir Nest