OpenAI LangChain Chat with Multiple PDFs Streamlit Web App

Python Streamlit web app allowing the user to upload multiple files and then utilizing the OpenAI API GPT 3.5 Turbo language models, the user is able to have a conversation about the uploaded documents. The user is also allowed to specify the language model and the temperature/creativity of the model. Also presented with a drop down for PDF analytics.

Example Screenshot V5

V5 Improvements

1. OpenAI API Authentication System acting as a login system
2. Improved Modularity and Readibility of Code

Example Demo V4 - multiPDF_V4_Demo.webm

Video Demo Coming Very Soon Improvements Include:

1. Improved Modularity and Readibility of Code
2. Added PDF Analytics Sidebar Dropdown
3. Implemented VADER Sentiment Scoring (Valence Aware Dictionary and sEntiment Reasoner) Visualizations
4. Added Tab to Display Source Code

Example Screenshot V3

1. Implemented Prompts for specific roles and personalities, currently: general assistant, academic, witty
2. Implemented a PDF DropDown for PDF Analytics - currently: word and character count
3. Improved modularity of code

Example Screenshot V2

1. Improved Error Handling
2. Added customizable settings (Model, Temperature)

Example Screenshot