Ride Dispatch Simulator

A real-time ride dispatch simulator built for a software engineering internship portfolio. It models driver supply, rider demand, ETA-based matching, surge pricing, live trip state, and dispatch latency.

Why this project fits an Uber SWE application

• Geospatial matching with Haversine distance and ETA scoring.
• Real-time state propagation over WebSockets.
• Supply-demand based surge pricing by local area.
• Trip state machine: waiting, matched, picking up, on trip, completed, cancelled.
• Load testing that reports throughput and p95 dispatch latency.
• Browser dashboard that visualizes drivers, riders, active trips, and surge pressure.

Tech stack

• JavaScript / Node.js
• Native HTTP server
• Native WebSocket protocol implementation
• HTML5 Canvas
• CSS Grid and responsive UI
• GitHub / Git

Benchmark

Measured locally with DRIVERS=10000 REQUESTS=10000 npm run load.

Metric	Result
Simulated drivers	10,000
Simulated ride requests	10,000
Matched requests	9,983
Match rate	99.83%
Throughput	770 requests/sec
p95 dispatch latency	0.12 ms
Average ETA	1 min

Run locally

npm start

Open http://localhost:3000.

Test and load test

npm test
npm run load

Optional larger run:

DRIVERS=10000 REQUESTS=10000 npm run load

Architecture

flowchart LR
  Browser["Dashboard"]
  HTTP["Node HTTP server"]
  WS["Native WebSocket stream"]
  Sim["Dispatch simulator"]
  Match["ETA + distance matching"]
  Surge["Local supply-demand surge"]

  Browser -->|"POST /api/request"| HTTP
  Browser -->|"GET /api/snapshot"| HTTP
  Browser <-->|"live snapshots"| WS
  HTTP --> Sim
  WS --> Sim
  Sim --> Match
  Sim --> Surge

Loading

Resume bullet

Built a real-time ride dispatch simulator with WebSocket telemetry, geospatial ETA-based matching, dynamic surge pricing, and load tests simulating thousands of drivers and ride requests while tracking p95 dispatch latency.

Future improvements

• Replace linear nearby-driver scans with Redis GEO or an S2/geohash index.
• Persist historical trip and dispatch metrics in Postgres.
• Add OpenTelemetry traces around dispatch scoring.
• Add replayable traffic scenarios for deterministic benchmark comparisons.