Upload any pdf file and ask questions from that pdf.
PaperBot generates embeddings from OpenAI embedding model and stores them in Pinecone Vector database. To answer user's questions, PaperBot does a semantic search on Pinecone Vector database and further refines the result with OpenAI completion model.
- OpenAI API Key
- Pinecone Database ENV and KEY
- MongoDB cluster username and password
- AWS S3 bucket, access key and access id
- Clone the repo
- Run
npm install
to install all the dependencies - Create a
.env.local
fromenv-example
file:cp env-example .env.local
- Update the OpenAI, Pinecone keys, MongoDB creds and AWS S3 bucket and creds in the
.env.local
file - Run
npm run dev
to start the project
- NextJS
- TailwindCSS
- MongoDB
- Pinecone (Vector DB)
- OpenAI Models - Embedding and Completion