After finishing a bookshop website project, an idea kept nagging at me: what if the site didn't just list books, but actually recommended the right one based on what someone was passionate about? That question is what pulled me into AI engineering.
Starting From an Itch, Not a Curriculum
Nobody assigned this project. It came from genuinely wanting to solve the gap I'd just seen — a bookshop site that could show books, but couldn't help someone who didn't already know what they wanted. I started learning AI engineering specifically to close that gap, which meant the learning curve had a real target instead of being abstract.
Building the Recommendation Layer
The core idea was simple to describe and harder to build well: take signals about a person's interests and match them to books that actually fit, instead of generic bestseller lists. Getting the recommendation logic to feel genuinely useful — not just technically functional — took more iteration than the rest of the project combined.
Why This Project Changed My Direction
This was the first time AI wasn't a buzzword I was reading about — it was a tool I'd used to solve a specific, personal problem I cared about. It reframed how I thought about my own skill set: not "web developer who might learn AI someday," but someone who reaches for AI whenever it's the right tool for a real problem. That reframe is still shaping the projects I choose today.