NoteFusion

• 

  landing

• 

  home

• 

Inspiration

Product managers and startup founders rely heavily on qualitative, not quantitative data, to make product decisions. Their primary touchpoints with customers is user interviews. However, user interviews are useless if each lives in a siloed document. How can we assist product managers and startup founders in create a coherent body of knowledge from their user interactions?

What it does

From a note, the users can choose a specific paragraph of note and query for semantically similar information across the database with one shortcut. The query sidebar will return not only the relevant notes, but also point out exactly which line/paragraph of these notes that are relevant to the query.

How we built it

• User interface: Next.js with Typescript
• Backend API: Python Flask
• Database: Pinecone (vector database) and Firebase

We store the original documents in Firebase Firestore, but we also dissect our notes into paragraphs, embed them into vectors, and store the vectors into Pinecone for the purpose of semantic search.

Challenges we ran into

Solidify the idea of the project was the most difficult part we ran into because there were various ways we could go about the solving the problem. This led in miscommunication in development that hindered our progress.

Accomplishments that we're proud of

Despite not having much experience in full-stack development or AI, we still completed the project and learned many valuable skills along the way.

What's next for NoteFusion

Given more time, we would like to enhance our product's feature to even give a detailed analysis on the content highlighted, basing on past notes, suggesting personalized approaches for notetakers.

This app uses generative AI and personalization to enhance your note-taking experience. It improves the quality of your notes by clarifying any gibberish and incorporates easy commands without relying on other sources like MLA, LaTeX, and more. This transforms your disorganized notes into easy-to-understand and personalized content, tailored to your learning needs.

Built With

• firebase
• flask
• nextjs
• pinecone
• python
• typescript