Senior Backend Architect Needed – Build a High-Performance Standalone Booking Engine

We are building a next-generation appointment booking SaaS for barbershops and salons. I am NOT looking for a frontend developer. I am NOT looking for someone to build an entire SaaS. I am looking for an experienced backend engineer or software architect to design and implement a standalone booking engine that will become the core scheduling service of our platform. This project focuses only on the booking engine. Scope The engine must support: Availability calculation Slot generation Temporary slot holds Appointment booking Appointment cancellation Appointment rescheduling Timezone support Staff working hours Breaks Time off / vacations Multiple services with different durations Multiple staff members Business booking rules Booking windows Lead time rules Buffer before/after appointments No double booking No overlapping appointments The engine should expose a clean REST API. Important This is NOT just CRUD. This is a scheduling engine. I expect clean architecture and deterministic behavior. Performance Requirements The engine should be designed for high performance. Target goals: Availability requests under 100–150 ms (typical workload) Booking under 200 ms Support concurrent booking requests safely Prevent race conditions Zero double bookings Database-level consistency Technical Stack Preferred: Node.js TypeScript NestJS (preferred but optional) PostgreSQL Prisma (optional) Redis (optional, only if justified) What I Expect I care much more about architecture than speed of development. I expect: Clean architecture Modular design Well-separated business logic Extensive automated tests Clear API design Scalable solution Deliverables The project should include: Standalone booking engine REST API Database schema Automated tests Documentation Docker setup Example requests To Apply Please answer the following: Have you built a scheduling or booking engine before? Have you worked with: Calendars Scheduling Reservation systems Medical booking Salon booking Hotel booking Airline booking How would you prevent double booking under heavy concurrency? How would you structure this project? What would you use: PostgreSQL only? PostgreSQL + Redis? Projection tables? Other techniques? Please share GitHub repositories or similar projects if available. Nice to Have Experience with: Booking engines Medical scheduling High-concurrency systems Calendar systems Reservation platforms Distributed systems Budget The user-facing booking flow must feel instant. Availability should load in under 300ms p95, and final booking confirmation should complete under 1 second p95 under normal workload. Open to discussion based on experience.