From full-stack to AI/ML

My earlier work is full-stack: typed Next.js applications, relational schema design, and deployed React front ends. Rather than treating that as unrelated history, I treat it as the foundation the AI/ML work builds on. Most candidates who come to machine learning through a purely academic path can produce a notebook; far fewer can take that model and wrap it in a real API, containerize it, and ship it behind a usable interface. That gap — between a result and a product — is exactly where a full-stack background becomes an advantage rather than a distraction.

The Django RAG engine in the Projects section reflects that directly: it is not a LangChain quickstart, but a system with a real backend, a vector store, containerized deployment, and multi-model support — architecture-level decisions, not just a call to an API.