Chat With RSS

A chat interface for interrogating uploaded files

Install

# 1. Create your virtual env.  Many tools for this, choose what you like best

# 2. With the virtual env activated, Install Python modules
(.venv) pip install -r requirements.txt

# 3. Configure
(.venv) cp dist.env .env 
# add your openai key

Initial ingest

This will ingest the most recent 100 items from the feed, create embeddings and persist into the Chroma vector db

# activate your python virtual env for this project
# run the ingest script.  
(.venv) python ingest.py
INFO:chat_with_rss.rss_reader:Found 100 entries in the feed.
...

This will take some time (2-3 mins) due to the fact it is downloading the content for (up to) 100 web pages.

You can run this repeatably and it will only download feed items not already processed

Run the app

After the install and the initial ingest, you can run the app locally

(.venv) streamlit run app.py

You can now view your Streamlit app in your browser.

Local URL: http://localhost:8501

Want to learn more?

My weekly newsletter will keep you informed on Generative AI topics important to software developers (without all the hype)