Production AI automation workflows. Real problems, real solutions, real results.
Built and maintained by Michelle Sartorio.
| # | Project | What it does | Status |
|---|---|---|---|
| 1 | Morning Briefing | Multi-source AI orchestration. Pulls weather, calendar, tasks, and news into one daily Telegram message. | 🟢 Live |
| 2 | Reel Transcriber Bot | Instagram reel transcription via Telegram bot. Send a link, get a transcript. | 🟢 Live |
🟢 Live = running daily in production
Every case study follows the same format:
1. Problem What manual task needed automating
2. Stack Tools and APIs used
3. Flow diagram Visual map of the automation
4. Iterations What changed from v1 to current
5. Failures What broke and how it was fixed
6. Result Measurable outcome
| Folder | Contents |
|---|---|
morning-briefing/ |
Full case study: problem, stack, flow, iterations, failures, result |
reel-transcriber/ |
Full case study: problem, stack, flow, iterations, failures, result |
These are not demos or tutorials. These are production automations that run every day. Each one was built iteratively: the first version broke, the prompt was wrong, the API response format changed, or the edge case nobody expected showed up.
The case studies document every failure and fix. That is the real value. Anyone can build a bot that works once. The hard part is building one that works every day.
Full write-ups with visuals live on the portfolio site.
