About

Han Cao is a highly motivated and skilled MSCS student at UCSD with a strong background in computer science. With a Bachelor of Engineering in Electronic Information Science and Technology from Tsinghua University, Han has gained valuable experience in the field of AI and software development. Han has 0.85 years of relevant experience, having previously worked as a Software Engineer in the Enterprise AI team at ByteDance. During his tenure at ByteDance, Han was responsible for the ranking of recommended apps for the advertising page of Xiaomi's Appstore. He accelerated the Spark daily Profile data generation by up to 3-4x by dynamic allocation, adaptive query execution, and slow nodes analysis. He also built users' App preferences features in the feature engineering library, which reduced feature size significantly and brought an increase in model AUC. In addition, he designed 20+ Apps' statistical features in the library, which served as biases and brought 0.06% increase in performance. Han also constructed user sequence features using Spark and Luigi, investigated different sequential models including DIN, and gained 0.2% increase in model performance. He flushed features to Redis for online serving and streaming training. He optimized the model through NAS-based feature selection, re-trained NAS-selected model from scratch, and achieved an increase in model performance. He reduced input size of Tower in the model by more than 30%. Currently, Han is working as a Data & Applied Scientist in the Bing Ads team User Modeling at Microsoft. He is responsible for developing machine learning models to improve the performance of Bing Ads. He is proficient in a variety of technologies, including Research Scientist, ML, Software Engineer, Spark, AI, and Redis. Han is interested in pursuing machine learning engineer and software development intern positions. With his strong technical background and experience in AI and software development, he is well-equipped to take on challenging roles in the field of computer science.