PetMatch

• 

  Landing Page

💡 Inspiration

Sequoia Humane Society shared a common challenge: too many pets, not enough personalized adoption. Their manual process for matching adopters and scheduling visits was inefficient, leading to longer stays for pets and missed connections with potential adopters. I was inspired to build a smarter system that could streamline operations and make adoption more meaningful.

🔍 What It Does

PetMatch uses a machine learning–based recommendation engine to pair adopters with compatible pets based on questionnaire responses. It offers:

• Real-time visit scheduling and confirmation
• Admin dashboard for managing pets and adoption requests
• Secure image uploads and automated email reminders
• A responsive, accessible user experience for all devices

🛠️ How I Built It

Frontend: React.js + Tailwind CSS Backend: FastAPI (Python) Database: PostgreSQL ML Matching Engine: Encodes adopter preferences and pet traits as vectors for similarity scoring APIs:

• OpenAI for pet summaries
• AWS S3 for image uploads
• MailerSend for emails Deployment: Vercel (Frontend) + Render (Backend & DB)

🚧 Challenges I Faced

• Building a robust vector-based matching algorithm that wasn't too complex
• Designing a secure and intuitive multi-role system for adopters and admins
• Managing async image uploads and email notifications reliably
• Aligning priorities while balancing scope

📚 What I Learned

• How to effectively prototype a full-stack ML-driven app
• The importance of clean UX for nonprofit workflows
• Real-world experience working with external APIs, user auth flows, and deployment pipelines
• How to stay focused, adaptive, and efficient when working solo on a high-impact project

Built With

• amazon-web-services
• fastapi
• mailersend
• openai
• postgresql
• python
• react.js
• render
• tailwind
• vercel