Grant Finder

Helping non-profits spend less time searching, more time creating impact.

Comment

Grant Finder

Helping non-profits spend less time searching, more time creating impact.

Inspiration

Non-profit organisations often operate with limited manpower but face a complex and fragmented funding landscape. Grants are published across multiple portals, each with different criteria, timelines, and reporting requirements. Through conversations with social service organisations, we learned that relevant grants are frequently discovered too late, not because they don’t exist, but because teams do not have the capacity to constantly monitor portals or interpret nuanced grant criteria while planning programmes.

We were inspired to build Grant Finder to help non-profits move from reactive grant searching to proactive, informed funding planning, so they can focus on delivering impact instead of navigating paperwork.

What it does

Grant Finder is an AI-powered grant discovery and matching platform that helps non-profits quickly identify grants that truly fit their mission and project needs.

It:

  • Automatically pulls and structures grant data from the OurSG Grants portal
  • Performs deep analysis on grant details, including linked pages and documents
  • Matches grants against an organisation’s mission, issue area, KPIs, funding quantum, and timelines
  • Ranks grants by relevance and urgency
  • Provides clear AI-generated reasoning explaining why each grant is a good fit

This reduces hours (or days) of manual research into a matter of minutes.

How we built it

Grant Finder is built using a modern, scalable stack:

  • Next.js for a fast, responsive web interface
  • Supabase & Python backend for data processing and business logic
  • Gemini AI to perform deep grant analysis, relevance scoring, and reasoning

At a high level, our system:

  1. Scrapes and structures grant data from the OurSG Grants portal
  2. Enriches each grant with deeper context by parsing linked pages and documents
  3. Uses AI to evaluate alignment between grants and user-provided organisational needs
  4. Outputs ranked, explainable grant recommendations

Challenges we ran into

  • Unstructured data: Grant details often reside in long text descriptions or external documents, requiring careful parsing and interpretation.
  • Nuanced eligibility criteria: Many grants involve implicit requirements that are difficult to capture with simple keyword matching.
  • Balancing depth and speed: Deep analysis improves accuracy but must still deliver results quickly.
  • Designing explainability: Making AI recommendations transparent and understandable to users was critical for trust.

Accomplishments that we're proud of

  • Built an end-to-end AI-powered grant intelligence pipeline within a hackathon timeframe
  • Successfully moved beyond keyword search to context-aware grant matching
  • Delivered explainable recommendations that help users make confident decisions
  • Created a solution that directly addresses real operational pain points faced by non-profits

What we learned

  • Grant discovery is as much about timing and context as it is about access to information
  • Non-profits need decision support, not just data retrieval
  • Explainability is essential when using AI in high-stakes funding decisions
  • Small improvements in workflow efficiency can have outsized impact in resource-constrained organisations

What's next for Grant Finder

Next, we aim to:

  • Add proactive alerts (through email) for newly published or recurring grants
  • Incorporate grant readiness and feasibility scoring
  • Expand beyond a single portal to include more funding sources

Grant Finder has the potential to become a core funding intelligence platform. Helping non-profits plan smarter, apply earlier, and sustain their impact over the long term.

Built With

Share this project:

Updates