mejaz / paperbot

Upload a PDF and just start asking questions from the PDF. It uses OpenAI's embedding model, chat completion model and pinecone vector database

PaperBot

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.

Pre-requisites to run this project

1. OpenAI API Key
2. Pinecone Database ENV and KEY
3. MongoDB cluster username and password
4. AWS S3 bucket, access key and access id

Steps to run this project

1. Clone the repo
2. Run npm install to install all the dependencies
3. Create a .env.local from env-example file: cp env-example .env.local
4. Update the OpenAI, Pinecone keys, MongoDB creds and AWS S3 bucket and creds in the .env.local file
5. Run npm run dev to start the project

Tech stack

• NextJS
• TailwindCSS
• MongoDB
• Pinecone (Vector DB)
• OpenAI Models - Embedding and Completion

Contact

[email protected]

paperbot's Issues

error from myfiles.js

components\MyFiles.js (9:10) @ map

7 | <div className={"bg-[#108dc7] text-primary-contrastText p-1 px-3"}>My Files
8 | {

9 | files.map((file, index) => (
| ^
10 | <button
11 | key={index}
12 | type="button"

Having issue