← InfraDraft

Geospatial

Design Uber / Lyft (Geospatial querying, matching systems)

A ride-hailing platform continuously ingesting driver GPS pings, indexing them geospatially, and matching riders to nearby drivers under tight latency budgets.

Open and Simulate this Architecture in InfraDraft

Core Architectural Components

Real-Time Location Ingestion Pipeline

Absorbs a continuous stream of driver GPS pings at massive volume.

Geospatial Index

Geohash, quadtree, or H3-based index enabling fast nearby-driver queries.

Rider-Driver Matching Service

Assigns the best available driver to a ride request under time pressure.

Dynamic/Surge Pricing Engine

Adjusts price in real time based on local supply/demand imbalance.

Trip State Machine

Tracks requested → accepted → in-progress → completed, driving all downstream side effects.

ETA & Routing Service

Computes route and arrival estimates for both rider and driver.

Payment & Fare Calculation Service

Computes and charges the final fare once a trip completes.

Driver/Rider Notification Service

Keeps both sides updated on match, arrival, and trip status.