EDGAR AI - chat with SEC filings

Simple web UI frontend that provides answers to free-text questions using SEC filing data as context.

Live demo: https://edgar-ai.streamlit.app/

Example usage:

- Question: What was Microsoft's revenue in fiscal year 2022?

- Sample response: Microsoft's revenue in fiscal year 2022 was $198,270 million.

- Question: How much R&D expense did TSLA have in each quarter of 2019?

- Sample response: The R&D expenses for TSLA in each quarter of 2019 were as follows:

- Q1 2019: $340.2 million

- Q2 2019: $323.9 million

- Q3 2019: $334 million. Please note that these figures are in millions of dollars.

Install

• Clone this repo git clone https://github.com/sputnik516/chatWithEDGAR.git

• Create a new conda env and install all libraries listed in requirements.txt:

  conda create -n edgar python=3.11.7 --file requirements.txt

  conda activate edgar

• Install all libraries listed in requirements.txt

• Create a file keys.py, by making a copy of keys_sample.py and renaming it. Set the OPEN_AI_KEY and KAYAI_API_KEY to OpenAI and Kay.ai API keys, respectively

• Launch the streamlit server by running:

  streamlit run streamlit_app.py

Hint: if running in Pycharm debugger, then set the "Run/Debug Configuration" to:

• Script Name: [path to streamlit executable, get it by running which streamlit]
• Script Parameters: run streamlit_app.py