RexVet: 24/7 Veterinary Telehealth Platform
Project Links
Technologies
About This Project
RexVet: is a full-stack veterinary telehealth platform that enables pet parents to book online consultations, join live video calls with licensed vets, receive prescriptions, and manage follow-up care through role-based dashboards. A major engineering focus was building a reliable **end-to-end consultation workflow**: 1. Pet parent books an appointment based on doctor availability. 2. Doctor joins secure video consultation. 3. Consultation audio/video is recorded in-session. 4. Recording is uploaded to **AWS S3** using a presigned-upload flow. 5. The system processes consultation context (recorded conversation + vet instructions + case details) to generate an **AI-powered veterinary health report**. 6. Final report is stored and available for review, continuity of care, and future visits. I also implemented **doctor-first scheduling automation** so doctors configure weekly availability only once (day + start/end time), and the system automatically generates appointment slots. This removed repetitive manual slot creation and reduced scheduling errors. The platform includes real-world operational modules: - Global timezone-aware appointment management for cross-region users. - Prescription and PDF generation workflows. - Payments and donation flows with Stripe. - Real-time messaging and notification systems. - Admin panel for appointments, doctors, pet parents, support operations, and analytics. - Secure authentication/authorization with role-based access for pet parents, vets, fosters, and admins. This project was built to handle production-like complexity: healthcare-style data handling, async media upload pipelines, AI-assisted documentation, and scalable API-driven architecture across a large feature surface.
Challenges Faced
1. **Global time-zone scheduling in real-world usage:** Pet parents and veterinarians were booking from different regions, so a major challenge was preventing time mismatch errors (wrong local time shown, missed appointments, daylight-saving changes). I implemented timezone-aware appointment handling so slots are stored consistently and displayed in each user’s local timezone.
2. **One-time doctor availability setup with automatic slot generation:** Doctors should not manually create slots every day. The challenge was building a system where a doctor sets weekly availability once (day + start/end time), and the platform automatically generates and maintains appointment slots from that profile configuration.
3. **Avoiding double-booking and slot conflicts:** In production, concurrent bookings can create race conditions. I had to enforce slot-state controls and validation so the same slot cannot be booked twice and manual/admin updates do not break schedule integrity.
4. **Balancing flexibility with operational simplicity:** Doctors needed easy control (set availability once), while admins needed visibility and override tools. The challenge was designing scheduling logic that is automated for doctors but still manageable for support/admin teams when exceptions happen.