Yuxiang Wang

About Yuxiang Wang

As a Data Scientist with a solid foundation in machine learning, underpinned by a strong mathematical background, I specialize in deep learning. My extensive hands-on experience, particularly in Natural Language Processing (NLP), allows me to transform complex datasets into actionable insights, leveraging cutting-edge technologies to solve critical business problems and drive innovation in data-driven decision-making. Throughout my career development, I have established a profound understanding of the complexities and nuances of machine learning and NLP. My approach combines theoretical knowledge with practical application, empowering me to design and deploy innovative solutions tailored to real-world challenges effectively.My current research interests are centered on the exciting and rapidly evolving field of Generative AI and Large Language Models. This passion drives my commitment to pursuing at the forefront of technological advancements, continually expanding my knowledge and skillsets in this area.Skills:• Programming: Python, SQL, Rust, Core Java• Deep Learning: PyTorch, TensorFlow• Cloud Platform: Microsoft Azure, Amazon AWS, Databricks• Database: MS SQL Server, PostgreSQL, MySQL, MongoDB, Access, Oracle • NLP Domains: Text analysis, NER (Name Entity Recognition), Sentiment analysis, NLU (Natural Language Understanding), NLG (Natural Language Generation), Generative AI

Experience

1. Senior Data Scientist

   AIG

   Apr 2025 — Present · Atlanta, GA, US

   Underwriting Assistant (UWA) LLM Development (Palantir Platform): Driving the devel-opment and deployment of the company-wide Underwriting Assistant (UWA) project, leveraging largelanguage models (LLMs) within the Palantir platform to automate data extraction from documentsfor underwriters• Document Processing Design & Implementation: Led the design and implementation of corecomponents, achieving document classification accuracy exceeding 98% and delivering multiplecritical downstream data extractors, including specialized financial document extraction• Generative AI Core Methodology: Implemented advanced Generative AI techniques, includ-ing RAG, prompt engineering, and LLM Agent design, to enhance the system’s ability toaccurately and automatically pull necessary data from unstructured documents• Cross-Functional Collaboration: Partnered closely with Palantir engineers for day-to-day oper-ations, ensuring the robust and scalable deployment of the LLM-powered UWA solution

Education

• Shanghai Jiao Tong University

  Bachelor's degree, Mechanical Engineering

• Stevens Institute of Technology

  Master of Science - MS, Business Intelligence & Analytics