AI module Parts Finder
Parts Finder: the client problem came first.
Customers knew what they needed in normal language. The catalogue wanted exact product and fitment logic.
Built from a real problem Parts Finder
The work behind the module.
I built a product-finder layer that reads messy customer intent against verified fitment data, then returns a recommendation the data can defend.
Technical products get a safer buying path. The customer can ask like a human, and support gets fewer cleanups from bad guesses.
It reads the source material and handles the repeatable work, then stops for review before judgement or money gets involved.
“A demo can look clever. The working system has to hold up when the week gets messy.”
Bring the messy version. I’ll tell you where I’d start.
The FOMO, the half-built experiment, the workflow that keeps nagging at you: bring that. We’ll shape the first build from there.