Cheul

I am a Software Engineer on the Upstage Data Team.

I'm interested in two questions:

  1. How can we correctly incentivize data contributors so we can have more "good" data?
  2. How can humans and AI work together to create better data?

Towards that, I am building a system that helps people better evaluate and guide LLMs.

I have also worked on applying RAG to build a chat engine for virtual idols, as well as a system for the automatic evaluation of insurance claims.

Before joining Upstage, I worked as an AI Researcher at Silvia Health, where I trained models to detect cognitive impairments from speech on smartphones.

Prior to that, I was a Master's student in the Interactive Computing Lab at KAIST. I worked on AdaptiveESM — an active learning approach for balanced and efficient collection of emotions in the everyday setting — and the K-EmoCon Dataset, a multimodal biometric sensor dataset for recognizing emotions in the wild.

Blog