Mi Yan, Ph.D. Email & Phone Number

Mountain View, California, Us

➼ Facilitated company-wide adoption of AutoML, guiding teams across different business units in translating business challenges into machine learning solutions, defining business impact, designing workflows, and proposing performance metrics, while expanding AI Marketplace (AIM) capabilities across various domains, including time-series forecasting, NLP, and recommendation systems.➼ Led the development of high-impact AI models for critical projects such as classification, traffic forecasting, and customer segmentation, resulting in significant improvements in operational efficiency, cost savings, and revenue growth.➼ Advanced Retrieval-Augmented Generation (RAG), driving the evaluation and fine-tuning of embedding and reranker models to improve search relevance and performance. Designed and implemented an LLM-powered irrelevant query classifier to reduce hallucination in RAG workflows. These efforts significantly enhanced the accuracy and efficiency of our retrieval systems.

Us

➼ Dynamic Pricing– Built machine-learning dynamic pricing models by performing customer segmentation with annual profit of Millions of dollars– Collaborated with cross-functional teams to identify business opportunities, define business metrics, build machine learning models, and interpret results– Developed and maintained production level OOP pipeline codes on Microsoft Azure➼ Production Optimization– Led the oil production optimization project leading to 4% uplift in production with annual profit of tens of Millions of dollars– Designed workflow, preprocessed data, and built XGBoost time-series models to predict monthly oil production– Published a paper and gave a talk "Machine Learning and the Internet of Things Enable Steam Flood Optimization for Improved Oil Production” in The 1st International Workshop on AIoT at KDD19 (arxiv.org/abs/1908.11319)

New York, New York, Us

– Built and evaluated performances of various cyber attack identification models

Massy , Ile-De-France, Fr

– Performed data engineering to preprocess 200 TB seismic data– Built velocity model, visualized and presented results in weekly client meeting– Mentored freshmen in their transition to industries, e.g. goal setting, technical preparation– Received Above & Beyond Technical Excellence Award

Doctor Of Philosophy - Phd, Physics

M.S. And B.S., Applied Physics