Python Automation with AI APIs
A hands-on session building three real automation scripts with the Anthropic and OpenAI APIs — email triage, document summarisation, and a Slack notifier. All code is reusable.
68 min
Video coming soon.
What we covered
Three real scripts, zero boilerplate padding. We started with the Anthropic Python SDK, called the API, then built toward something actually useful.
Scripts we built
- Email triage bot — reads a Gmail label, classifies by urgency, drafts a reply for the high-priority ones
- Document summariser — takes a folder of PDFs, extracts text, returns a one-paragraph summary per document
- Slack notifier — posts a daily digest to a Slack channel based on a data source of your choice
Prerequisites
Python 3.11+, a text editor, and an API key from Anthropic or OpenAI. Everything else is explained live.
Why this matters
Most "AI automation" tutorials stop at a single API call and skip the parts that actually decide whether the script survives contact with a real inbox: error handling when the model returns something unexpected, rate limits, and where to draw the line between "let the model decide" and "hardcode the logic." Each of the three scripts in this session is built with that line explicit, so you leave with automations you can trust unattended, not just demo well.
Format
Recorded, 68 minutes, hands-on coding throughout — pause and follow along in your own editor. The GitHub repo and Colab notebook (linked above) mirror exactly what's built on screen.
Stay in the loop
Next session in your inbox.
I run sessions every couple of weeks. Drop your email and I'll send you the link when the next one is live.