This repository documents my hands-on learning journey through AI engineering — following structured coursework, practicing concepts, and building a finished project along the way. It's a working record of that process, not a single polished deliverable, so you'll find a mix of course exercises, practice notebooks, and one cleaned-up finished project.
.
├── barrier/ # Finished project — start here
├── project/ # Practice notebooks applying course concepts to the barrier content
└── course/ # Exercises following the course material directly
A deployed, cleaned-up RAG FAQ assistant for the debauchee/barrier GitHub repository. This is the polished output of the learning process below — documented, debugged, and running live on Streamlit Community Cloud.
See barrier/README.md for full details, setup instructions, and architecture.
Jupyter notebooks (.ipynb) showing the practice work that led to barrier/ — experimenting with the same underlying content (the debauchee/barrier repo) before consolidating what worked into the final project. This folder captures the messier, in-progress side of getting there: trying things, hitting issues, and figuring out an approach.
Notebooks (.ipynb) following along directly with the AI Hero/AI Agents Crash Course material (DataTalks.Club's free Data Engineering Zoomcamp FAQs). These are closer to structured exercises than original work, kept here as a reference of concepts covered.
course/ and project/ are intentionally left as-is for now — notebooks, exploratory code, and all. They reflect a genuine learning process rather than a finished product, and I'd rather be upfront about that than tidy them into something they're not. I may clean these up (or convert relevant pieces to .py scripts) in the future, but for now they're kept as an honest record of how barrier/ came to be.
If you're looking for the actual working project, head to barrier/.