Saehoon Kim

Soombit.ai. Pangyo, South Korea.

shkim.jpg

I am a co-founder and CTO of Soombit.ai. Previously, I led the research and productization of Karlo, an image generative model at Kakao Brain. Before that, I headed the machine learning team at AITRICS.

I hold a Ph.D. and a B.S. in Computer Science and Engineering from POSTECH, with additional internship experience at Microsoft Research Asia (MSRA) and Microsoft Research (MSR). During my Ph.D., I worked under the supervision of Prof. Seungjin Choi on research projects focused on randomized and learning algorithms for approximate nearest neighbor search.

My primary research interests include developing a chest X-ray preliminary report generation system and efficiently training large-scale generative models, particularly those based on autoregressive and diffusion models.

news

Nov 05, 2024 We are excited to share that we have secured KRW 5 billion in seed funding from Altos Ventures! This funding will enable us to greatly enhance the early version of our preliminary report generation system for chest X-rays. A big thank you to Altos Ventures for their support!
Aug 23, 2024 After several incredible years at Kakao Brain, I’m excited to begin a new chapter as Co-Founder and CTO at Soombit.ai. Our mission is to develop a chest X-ray preliminary report generation system, with a strong focus on integrating this solution seamlessly into workflow.
Feb 22, 2024 We’re excited to release the improved version of our text-to-image generation model, Karlo 2.1. Explore this model at karlo.ai and access the API through our developer site.
Jul 10, 2023 We’re happy to release Karlo 2.0, which is available at our developer site.
Dec 20, 2022 Karlo-v1-alpha is now available at Huggingface diffusers library.