Granted

Inspiration

After catching up with his friend over lunch, one of our teammates learned that the majority of his friend's summer was spent not divulging in the world of art history during her prestigious research internship but rather bent over a desk drafting grant proposals. The current grant-writing pipeline is fragmented, slow, and frankly, demoralizing. NGOs and labs spend billions of hours annually just on this administrative burden, taking away from time that could instead producing life-changing discoveries.

What it does

Our application brings everything about grant writing into one simple and connected space. The platform allows users to organize all their materials in one place, from drafts and notes to funding requirements and past submissions. It uses several AI models depending on the task, like ChatGPT for rewriting text or others for analyzing structure and compliance. Granted remembers previous chats and documents through a backend memory system, so the assistant can build on past conversations instead of starting from scratch. When someone highlights a section, the AI can quickly rephrase it to make it clearer or more polished, keeping the tone and meaning consistent. It also follows the specific formatting and rules tied to each grant, which makes switching between different applications easy and error-free.

How we built it

Our team comes from a diverse array of skills and experiences, which enabled us to divide the work and collaborate efficiently. After aligning on the concept, we began by setting up a robust database to manage users, documents, and contextual data. From there, we developed a document parsing system to extract structure and data from uploaded PDFs. Once this backbone was in place, we implemented a PDF upload and management layer to make ingestion seamless for users. Next, we integrated user authentication and authorization, ensuring that each user’s workspace remained secure and isolated. We then built the heart of the system: a chatbot interface that allows users to provide context, receive tailored writing suggestions, and iterate on drafts directly within the editor. Finally, we designed a landing page and polished UI/UX that made the workflow more intuitive.

Challenges we ran into

One of the biggest challenges we faced was connecting our database to the AI assistant in a way that allowed it to read and remember user context. Two of us were focused on the backend while others were working on the assistant logic, so getting those pieces to communicate smoothly took a lot of troubleshooting and late-night debugging. Since it was our first hackathon, figuring out how to collaborate efficiently on GitHub while also managing time effectively proved to be a steep learning curve.

Accomplishments that we're proud of

Our team combined different languages, frameworks, and AI models into a unified system that works end-to-end, connecting the backend memory database with the assistant to create a responsive and contextual writing experience. We built an MVP that not only demonstrates technical feasibility but also delivers real value by saving time and preserving knowledge for nonprofit teams. Beyond the code, we proved that under pressure we could plan, adapt, and execute as one unit.

What we learned

We learned that building something meaningful takes more than skill: it takes communication, trust, and a mutual dedication to the idea. Talking through problems, challenging each other’s assumptions, and learning how to make fast decisions together turned out to be just as important as writing any line of code. On the technical side, we learned how to connect multiple AI models into a single system, design a memory layer that lets the assistant retain context, and build an end-to-end workflow that actually feels usable.

What's next for Granted

Granted has the potential to evolve from a writing assistant into a full-fledged grant intelligence platform. Future development could focus on deepening AI contextual understanding by integrating real grant databases and learning from successful proposals, as well as adding collaborative features for teams and institutions. Automating administrative tasks like budgeting, compliance checks, and deadline tracking would streamline the workflow even further. Moreover, long term growth lies in offering data-driven insights into funding landscapes and success patterns for NGOs and researchers.