Lately I've been messing around with a personal chatbot that can answer questions about me, my work experience, projects I've built, and even how I usually learn new tech.
Instead of just showing a static portfolio, I thought it'd be way more fun if people could literally chat with an AI version of me and ask stuff like:
- Who's Wahyu?
- What projects has he worked on?
- Why is he into AI?
- How does he learn new technologies?
To make that happen, I'm using something called Retrieval Augmented Generation (RAG).
In simple terms, the AI isn't just relying on its own knowledge. It also reads from a custom knowledge base that I put together myself. That knowledge base contains my work experience, project stories, personal thoughts, notes, and other things that help describe who I am beyond what's written on a resume.
The flow is pretty straightforward:
Someone asks a question → the system looks for relevant information → that context gets sent to the AI → the AI generates an answer based on the data it found.
Because of that, the responses feel a lot more personal and grounded in real information, instead of random guesses or generic AI answers.
It's still very much an experiment right now, but the end goal is to build a "Digital Wahyu" that can explain who I am, what I've worked on, and how I approach solving problems—basically a chat version of me.