Wei Yang

Developed a Generative AI platform for end-user experiences, built LLM-based agents translating natural language and Greenplum SQL to Teradata SQL, and open-sourced an LLM attribution library to improve interpretability and deployment velocity.

Developed time-series change point detection and clustering algorithms, built Snowflake/Databricks ETL pipelines and Tableau dashboards to automate analytics and expose KPIs for stakeholders.

Built deep learning models to predict optimal read voltages and drive failure rates, developed recursive algorithms to reduce read failure triggers, and deployed models onto chip-level frameworks.

Improved embedding search performance using ElasticSearch, Faiss and product quantization, redesigned features with Spark, and constructed behavior graphs and user clusters using MinHash and neo4j for product analytics.

Implemented IRT and knowledge-graph embedding models with TensorFlow, and used Hive SQL to analyze e-learning recommendation effectiveness for large-scale educational datasets.

Doctor of Philosophy in Machine Learning with extensive research and applied expertise in deep learning, probabilistic modeling, and big data systems.

Master of Science in Statistics focusing on statistical modeling and high-dimensional data analysis to support machine learning applications.

Master of Science degree from Columbia University with advanced coursework relevant to machine learning and data science.

Bachelor of Engineering providing foundational engineering and computational training applicable to machine learning and systems work.