PDF Reading Chatbot

Initial setup

Install python and uv for "Extremely fast Python package installer and resolver, written in Rust"

brew install python
brew install uv
brew install poppler

Install dependencies

uv sync

Create a local .env file

You will need set up a number of API keys for this to work. OpenAI and Mistral AI require you to set up a payment plan and add payment credit for the API keys to work. Once you have the API keys, on your terminal run the below command

export OPENAI_API_KEY="YOUR_OPENAI_API_KEY_HERE"
export LANGCHAIN_API_KEY="YOUR_LANGCHAIN_API_KEY"

And in your .env file, add the below env vars

OPENAI_API_KEY="YOUR_OPENAI_API_KEY"
LANGCHAIN_API_KEY="YOUR_LANGCHAIN_API_KEY"
MISTRALAI_API_KEY="YOUR_MISTRALAI_API_KEY"
DATASET_PATH="/Users/{YOUR_USERNAME}/{PROJECT_DIR}/model/data"
DATASTORE_DIR="/Users/{YOUR_USERNAME}/{PROJECT_DIR}/PersistentDB"
LANGCHAIN_ENDPOINT="https://api.smith.langchain.com"
LANGCHAIN_TRACING_V2="true"

• Dataset path is the path to the dataset folder eg. "/Users/kimstocker/projects/pdf_reading_chatbot/model/data"
• Datastore is the PersistentDB path eg."/Users/kimstocker/projects/pdf_reading_chatbot/model/PersistentDB"

Run app

uv run streamlit run app.py --server.port=8510

A browser window should open with the app. Ask a question about NRMA and ANZ insurance policy. Summary table will show on top of the page

App images