Junho Song believes the best data is even more compelling than the best algorithm. As a deep learning data scientist, he actively participated in the top tier conference NIPS2016, CVPR2016, NIPS2017 and is working with deep learning researchers from all over the world. He is a Ph.D. Candidate at Seoul National University, and received his M.Sc. degree from KAIST. With many years of experience in applied practical deep learning, as a core planner and project leader, he has led two projects; DL Pipeline and Generating Training Dataset in military defense domain for 'Agency for Defense Development(ADD, S.Korea)'. Moreover, he is not only a serial entrepreneur with experience in founding two startups, but also has been leading one of the biggest deep learning open communities supported by S.Korean government and NAVER for many years.