Ooh Ahh
On-Demand Beauty Service Platform
A location-based service marketplace connecting users with nearby beauty professionals, similar to an Uber-style booking system for salon and home services.
System Architecture & Design Patterns
Modular Monolith with Event-Driven Socket Cluster & Redis Pub/Sub
Chosen to maintain high development velocity while allowing real-time websocket connections to scale horizontally across multiple instances without state locking or session stickiness issues.
- MongoDB 2DSphere indexing for sub-10ms geospatial provider discovery queries
- Redis Pub/Sub adapter syncing real-time job bidding & live provider coordinates across Socket nodes
- Stripe Webhook processing engine with idempotent event verification
Engineering Challenges & Solutions
High-frequency provider GPS updates created excessive backend CPU utilization and MongoDB write contention.
Implemented client-side distance thresholding combined with Redis in-memory location caching, reducing database write operations by 70%+ while preserving live map accuracy.
Payment webhook race conditions causing duplicate job booking confirmations.
Architected an idempotent transaction pipeline using Redis distributed locks (Redlock) and DB transaction state checks, guaranteeing single-execution semantics for payment events.
Tech Stack
Core Capabilities
- Location-based service discovery
- Real-time booking & provider matching
- In-app chat between user and provider
- Job bidding / acceptance system
- Live job status tracking
- Notification system (SMS/Push)