Learning full-stack engineering so I can orchestrate agents well and understand the whole process of creating with AI — not just consume it.
AI can be leveraged for almost anything. It is only as effective as the person using it. People will be consumers or creators. I am choosing the creator side.
At this point here, I am learning how to embed images. And here below is my test image:
Selected work
- collaborative-partner — Socratic system that helps you drill down to the right question. Can run with an AI attached for a more natural conversation, or without.
- NAPS2_filetree — Fork of open-source NAPS2. Split a scan batch into multiple documents and save multiple outputs. Other tweaks; still in progress.
- ODL PDF Extractor — Local extractor for scanned PDFs using OpenDataLoader. Private repo.
- PDF-Coordinates-Extractor — Draw a box on a PDF, get the coordinates. First repo I vibe-coded.
More on GitHub — t1ms.
Studying
ICM (folder structure as agent architecture) — project by Jake van Clief. I use icm-architect a lot; I am still digesting the scope of his research. Not my invention.