Design a Proximity / Ride-Sharing Service (Uber / Lyft).

In Simple Words: Connect drivers with nearby riders in real-time while processing location updates every 4 seconds from millions of active vehicles with sub-100ms spatial matching.

Spatial Indexing Technologies:

• Geohash: Hierarchical spatial data structure that subdivides Earth into bounding grid boxes and encodes them into Base32 strings. Nearby locations share common string prefixes!
  • Example: Prefix 9q8yy covers an area of roughly $4.9\text{ km} \times 4.9\text{ km}$.
• Google S2 / Uber H3: Uses hexagonal cells rather than rectangular boxes, eliminating corner distortion and ensuring all neighbors are equidistant.

flowchart TD
    Driver["Driver Phone App"] -->|"GPS Update Every 4s"| LB["Geo-DNS / Envoy LB"]
    LB --> LocationSvc["Location Ingestion Service"]
    LocationSvc --> RedisGeo["Redis Geospatial Engine (GEOADD / GEORADIUS)"]
    
    Rider["Rider App"] -->|"Find Rides Near (Lat, Lon)"| MatchSvc["Ride Dispatcher Engine"]
    MatchSvc --> RedisGeo
    RedisGeo -.->|"Return Top-10 Drivers within 3km"| MatchSvc
    MatchSvc --> DispatchSvc["Optimal Routing & Notification"]