All projects
Full-StackAI

DonorDash

Donation platform that matches donors with nearby charities

Problem

Donors don't know which nearby charity needs what they have.

What I built

A React + Express + MySQL platform with a location-based charity-matching model.

Result

List an item, get matched with a nearby charity, schedule pickup, see proof of delivery.

DonorDash preview
Role
Full-stack development
Year
2025
Stack
React, TypeScript, Vite, Tailwind CSS, Express, MySQL, scikit-learn, Google Maps API

Overview

A platform where people list items they want to give, get matched with verified charities that need them, schedule a pickup and receive proof of delivery.

The problem

People want to donate goods but do not know which charity needs what, and charities cannot tell donors what actually happened to their donation.

What I built

  1. 1Designed the donor journey in four steps: list items, smart matching, hassle-free pickup and delivery proof.
  2. 2Built the interface in React and TypeScript with Tailwind CSS, shadcn/ui and Google Maps for charity locations.
  3. 3Built an Express API on MySQL for sign-up, charity listings and donations with up to 10 photos per item.
  4. 4Trained a matching model in Python: K-means groups donors by location, and a Random Forest predicts the best-fit charity for each group.

Highlights

  • Location-aware charity matching from a K-means + Random Forest model.
  • Closes the loop with delivery confirmation for every donation.

How it works

  1. React + Vite client
  2. Express API + photo uploads
  3. MySQL
  4. K-means donor clusters
  5. Random Forest charity match

Screens

DonorDash landing page
Landing page and the four-step donor journey
Next project
Zoom-iT
City-wise food discovery and ordering