Servyq

On-demand home services frequently suffer from poor coordination between customers and service providers. Customers face unpredictable arrival times and opaque pricing, while technicians struggle with manual job matching and delayed disbursements.

The goal of Servyq was to engineer a unified two-sided marketplace that handles real-time job dispatch, technician location streaming, in-app messaging, and automated milestone payouts.

I led the end-to-end architecture and implementation: designing the React Native mobile application for both client and provider roles, implementing the NestJS microservices backend, and modeling the PostgreSQL database schemas.

Key constraints included supporting low-bandwidth network environments in emerging markets, ensuring zero-downtime ledger consistency during payment capture, and maintaining low-latency location broadcasting across mobile clients.

The backend is built around an event-driven NestJS architecture with Redis Pub/Sub managing live WebSocket connections for real-time technician tracking and status transitions.

Database transactions are orchestrated through PostgreSQL with strict isolation levels to prevent double-booking and payment concurrency anomalies. The React Native mobile frontend utilizes offline-first caching to ensure smooth operation when moving between coverage dead zones.

Dual-Role Mobile Client: Single codebase powering customer booking flows and provider job management with dynamic role switching.

Real-Time Dispatch Engine: Geo-spatial proximity matching algorithms assigning jobs to the closest available technicians within 300ms.

Escrow Settlement: Automated payment capture, milestone holding, and instant payout disbursement upon digital job sign-off.

Servyq was successfully shipped and deployed to production. The platform maintains sub-second dispatch latency and has processed thousands of live service bookings with zero ledger mismatch incidents.