Granted

Hack4Good 2026 - 2nd Place Winner

Granted won 2nd place at Hack4Good 2026! Granted is an automated grant discovery engine built to help NPOs find and act on funding opportunities.

Hackathon Project — Overview

This repository contains the Hackathon entry for "Granted": an automated grant discovery engine built to help NPOs find and act on funding opportunities.

Inspiration

We wanted to build a system where funding finds the NPO, shifting the paradigm from "Pull" (active checking) to "Push" (proactive notification).

Problem Statement

How might non-profit organisations "pull" information about grants from the OurSG grants portal that are relevant to them according to key criteria including issue area, scope of grant, KPIs, funding quantum, application due date, etc., so that they can strengthen their financial sustainability?

What it does

• Grant Discovery: Browse and search through available grants from OurSG with real-time filtering by category, eligibility, and funding amount.
• AI-Powered Assistant: Chat with an intelligent assistant to get help finding relevant grants, understanding requirements, and refining your search.
• Smart Matching: Automated matching engine that analyzes NPO profiles against grant criteria using vector embeddings and intelligent scoring.
• Saved Grants: Bookmark grants of interest and track them in a personalized dashboard.
• Email Notifications: Receive automated email digests with matched grants, including grant details, deadlines, and application links.
• NPO Profile Management: Create and manage organization profiles with mission, budget, focus areas, and KPIs for better matching.

Tech Stack

• Frontend: Vite + React + TypeScript + Tailwind CSS + Vercel
• Backend: Firebase Cloud Functions + Python SDK
• Database: Firestore (NoSQL; storing NPO, Grants, & Match Information)
• AI/LLM: OpenAI GPT-4o for chat interactions and grant matching
• Authentication: Firebase Auth

Challenges

• Data unstructuredness: transforming conversational text into structured fields (e.g., "Average Project Budget") required iterative LLM tuning.
• Balancing relevance vs. noise: refined scoring logic to avoid spamming users and focus on high-confidence matches.
• Real-time reliability: building a robust delta-detection mechanism for the OurSG portal with minimal false positives.

Accomplishments

• Built a push-based automation where new grants can trigger tailored emails to matched NPOs.
• Implemented a deadline-proximity boost to reduce missed deadlines.
• Replaced generic forms with an interactive conversational UX that produces structured profiles.

What's next

• Automated drafting: help NPOs generate first-draft proposals from stored profiles.
• Multi-source aggregation: expand monitoring to corporate and private philanthropy.
• Feedback loop: allow NPOs to rate matches to refine vector embeddings and recommendation quality.

---

Developer Guide (Quick)

This project combines a React + TypeScript frontend and a Python backend (serverless functions included). Use the sections below to run the project locally.

Quick Overview

• Frontend: frontend/ (Vite + React + TypeScript)
• Backend: backend/ (Python)
• Cloud Functions: backend/functions/

Prerequisites

• Node.js (LTS recommended) — verify with node -v
• npm (comes with Node) — verify with npm -v
• Python 3.10+ — verify with python3 --version
• Git — verify with git --version

Setup (repo root)

Clone and open the repository:

git clone https://github.com/ClarenceChoo/Granted.git
cd Granted

Frontend — Run locally

1. Change into the frontend folder and install dependencies:

cd frontend
npm install

2. Start the dev server:

npm run dev

3. Open the local URL shown in the terminal (default: http://localhost:5173).

Common commands (from frontend/):

• npm run dev — start dev server
• npm run build — build production bundle
• npm run lint — run linters

Backend — Run locally (Python)

1. Create and activate a virtual environment from the repo root (recommended):

python3 -m venv .venv
source .venv/bin/activate

2. Install backend requirements:

pip install -r backend/requirements.txt

3. Run the backend server (adjust to your runner):

python backend/main.py

If your backend uses uvicorn/gunicorn/flask adapt the command accordingly.

Email Notification System

The project uses Firebase's firestore-send-email extension to automatically send grant notifications to NPOs:

How it works:

1. When matches are created, the system adds email documents to the Firestore mail collection
2. The Firebase extension monitors this collection and sends emails automatically
3. Email templates are stored in backend/functions/template/ and rendered with grant data

Email Features:

• Mobile-responsive HTML templates with grant details
• Personalized recipient information
• Direct links to grant details and application pages
• Automatic tracking of email delivery status

Configuration:

• Extension settings: backend/extensions/firestore-send-email.env
• Email templates: backend/functions/template/email_template_updated.html
• Test emails: Use the EmailTester component in the frontend (frontend/src/components/features/email/EmailTester.tsx)

Handlers:

• send_email.py - Manual email sending endpoint
• send_grant_emails.py - Automated grant notification emails

Cloud Functions

Serverless handlers live in backend/functions/. See backend/functions/requirements.txt and backend/functions/main.py for details.

Project Structure

Granted/
├── frontend/
│   ├── src/
│   │   ├── components/
│   │   │   ├── features/
│   │   │   │   ├── chat/
│   │   │   │   ├── email/
│   │   │   │   └── grants/
│   │   │   └── layout/
│   │   ├── contexts/
│   │   ├── pages/
│   │   ├── services/
│   │   ├── utils/
│   │   ├── App.tsx
│   │   ├── firebase.ts
│   │   └── types.ts
│   ├── public/
│   ├── package.json
│   ├── vite.config.ts
│   ├── vercel.json
│   └── eslint.config.js
├── backend/
│   ├── functions/
│   │   ├── handlers/
│   │   │   ├── ai/
│   │   │   ├── grants/
│   │   │   ├── matching/
│   │   │   ├── npo/
│   │   │   └── saved_grants/
│   │   ├── services/
│   │   ├── template/
│   │   ├── tests/
│   │   ├── utils/
│   │   ├── main.py
│   │   ├── requirements.txt
│   │   └── pytest.ini
│   ├── extensions/
│   ├── firebase.json
│   ├── firestore.rules
│   ├── firestore.indexes.json
│   ├── pyproject.toml
│   └── sample.env
├── assets/
└── README.md

Notes & Troubleshooting

• If ports are occupied, the dev servers will usually try a different port — check the terminal output.
• If you hit dependency issues, recreate the environment (.venv) and reinstall packages.
• For frontend build issues, delete node_modules and package-lock.json, then run npm install.

Contributing

• Open an issue for new feature requests or bugs.
• Create small, focused pull requests.

Connect with Us

• Anson Ng (Frontend Engineer/Product Manager) - linkedin
• Clarence Choo (Frontend Engineer/UIUX Designer) - linkedin
• Tan Yu Hoe (Backend Engineer/Database Administrator) - linkedin
• Lionel Zheng (Material Scientist) - linkedin