Scroll
Selected Work
Designing
Intelligence
Building AI products isn't just about wrapping an API in a chat UI. It requires a fundamental shift in how we approach trust, state, and user agency.
Case Study Highlight
Neuro Prompt Interface
Read how I approached conversational UX and parameter steering for a complex LLM agent dashboard.
Transparent Intent
AI shouldn't pretend to be human. Interfaces must clearly communicate when a system is generating content versus fetching facts.
Graceful Fallbacks
LLMs hallucinate and APIs timeout. Designing robust error states and human-in-the-loop overrides is non-negotiable.
Augmentation Over Automation
The goal is to elevate human capabilities, not replace them. We build copilots that give users agency and steerability.