I make ambitious systems easier to see.

I build inspectable software systems that turn ambitious AI and Web3 ideas into working product flows, then document the decisions and evidence behind them.

I am a Software Engineering student at Telkom University who learns by building systems with real constraints. My favorite projects sit where product behavior and technical architecture have to agree: an AI agent that should help without taking custody, a payment flow that needs a receipt, or a Web3 action that should be understandable before someone signs it.

I usually work across the stack, but I do not treat full-stack work as a claim that one person did everything. The strongest hackathon projects here were team efforts. I like leading the product direction, connecting specialist contributions, and making the final path coherent enough for a user, teammate, or reviewer to inspect.

That working habit also shapes my research into LLM-assisted self-healing test automation. Before asking an intelligent system to act, I want the failure, decision boundary, and recovery evidence to be visible. The same principle appears in Fradium, Nova, PayGate, and Quorum in different forms.

This portfolio is a living record of that practice. Some projects are hackathon artifacts, some are research, and PayGate is an active product build. I present each one with its actual state, the role I held, the people around it, and the proof that survives after the demo.