Selected work
DataGPT AI
Product engineering for a data analytics platform where workflow clarity, query-state visibility, and trust in system behavior mattered as much as raw capability.
- Built interfaces that made complex query and result states easier for users to follow.
- Designed orchestration UI that surfaced pipeline state, system behavior, and failure conditions clearly.
- Focused analytics UX on clarity and trust instead of dashboard decoration.
What it is
DataGPT is a data analytics platform aimed at making complex data exploration easier through AI-assisted query workflows.
My work focused on turning orchestration and analysis into an interface users could actually follow. That meant showing how a request moved through the system, what state it was in, when it failed, and what the user could do next.
Why it matters
In analytics products, trust breaks when users cannot tell what the system is doing. If result generation feels opaque, recovery is unclear, or failures are hidden behind generic loading states, the product quickly starts to feel unreliable.
This work treated visibility into system behavior as a core product requirement, not just a technical detail.
Technical shape
The core challenge was making complex workflow state legible through the UI. That required strong state handling, clear interaction design, and visualization choices that made AI-assisted analysis feel understandable instead of black-boxed.
The point was not just to return results. It was to make users trust the path that produced them.