Why I Built This Site as an AI-First Portfolio
A portfolio should feel like a product. Static pages are for resumes; sites should talk back.
A resume tells you what someone did. A good portfolio site should let you explore how they think. This one does both, and it talks back.
The shape
The site is a Next.js app with a Gemini-grounded assistant, two interactive visualizations — one for cloud infrastructure, one for the LLM pipeline — and a small corpus of case studies, projects, and system design deep dives. Everything is built in public so the assistant has something honest to ground on.
Why an assistant
Because most people reading this site have a specific question: "Could Hassan build this? Could he help with that?" Static content answers those questions slowly. The assistant is grounded in the same case studies you'd read manually, and it will happily compare trade-offs or walk through an architecture decision with you.
Why the visualizations
Because architecture is hard to describe in prose. The cloud topology shows a reference backend — API, events, data, workers — and the LLM pipeline shows what actually happens between a question and an answer. Both are built in plain SVG with Framer Motion, so they're fast and accessible.
What's next
More writing. More visualizations. And a RAG layer over the structured content so the assistant gets sharper as the corpus grows.