Professional Project
Lyfsum
Overview
Lyfsum is a healthcare appointment booking platform connecting patients with medical providers. I built the full-stack system — patient search and discovery, doctor availability, and appointment management — for a healthcare startup aiming to simplify how people access medical care.
Role
Full Stack Developer at Designare Solutions
Problem
Healthcare appointment scheduling is fragmented: patients call multiple clinics, face hold times, and still end up with inconvenient slots. The startup needed a unified platform where patients could search for providers by specialty, view real-time availability, and confirm bookings in a single flow — without requiring heavy infrastructure investment.
Solution
Built a React web app with provider search, filtering by specialty and location, and a calendar-based slot picker. The Node.js backend handled appointment state management (pending, confirmed, cancelled, completed) with PostgreSQL as the relational store. Provider availability was modeled as recurring and exception-based windows to support real clinical scheduling patterns.
Architecture
A React SPA for the patient-facing booking experience, a Node.js REST API for appointment and provider management, and PostgreSQL for relational healthcare data. Deployed on GCP with Cloud SQL for managed database infrastructure.
Key Design Decisions
- Provider search with specialty, location, and availability filters using parameterised SQL queries
- Slot availability engine supporting recurring schedules with override/exception windows
- Appointment state machine (requested → confirmed → completed / cancelled) enforced at the API layer
- Role-aware API: patients and providers have distinct access scopes via JWT claims
- Email and in-app notification system for appointment reminders and status updates
- PostgreSQL on GCP Cloud SQL for managed, HIPAA-considerate data storage
Challenges
- Modeling provider availability windows that support recurring schedules, exceptions, and real-time slot blocking
- Preventing double-booking under concurrent appointment requests for the same slot
- Designing a data model that could accommodate multiple provider types without schema migrations per type
- Keeping the patient search experience fast with compound availability + specialty filtering
Impact
- Delivered a production-ready healthcare booking platform from design to deployment
- Eliminated phone-based scheduling friction for patients through a self-serve booking flow
- Supported multiple provider types and specialties under a unified data model with no schema rework