Peter Lu

About Peter Lu

Seasoned AI leader with over 10 years of experience in generative AI and deep learning, leading data science teams to define strategic roadmaps, identify high-impact projects, develop and drive adoption of AI technologies. Proven track record of delivering end-to-end enterprise solutions that have generated over $60M in business value- Deep expertise in neural network architectures for quantitative analysis, computer vision & large language models (LLMs) • TradingAgents: multi-agents LLM financial trading framework • GANs, VAEs, diffusion models, and transformer-based architectures for LLMs • RNNs and reinforcement learning for sequential modeling and decision-making • 2D/3D encoder-decoder frameworks for object detection and segmentation- Strong mathematical foundation in machine learning and deep learning algorithm development- Extensive hands-on experience in coding with TensorFlow & PyTorch and deploying scalable AI solutions on large-scale GPU - Broad software engineering background with DevOps for building robust, production-ready AI systems- Effective team management and communication skills with a history of cross-functional collaboration and technical leadership

Experience

1. Senior Manager, Data Science

   Hess Corporation

   Nov 2019 — Present · Houston, TX, US

   Lead a data science team in the development of AI software and intelligent agents, significantly boosting productivity and generating measurable profit gains- Defined strategic roadmap and championed enterprise-wide excellence in data science- Conducted cutting-edge research and evaluated state-of-the-art deep learning algorithms for generative AI applications- Deployed deep learning models on cloud delivering scalable AI solutions across the organization- Provided technical leadership through code reviews, architecture planning, and mentorship- Hands-on expertise in Python, TensorFlow, and PyTorch for the following implementations: • AI-powered trading platforms with fundamental, sentiment, news and technical analysis • AI agents to reduce operational risk and minimize downtime by analyzing geological and drilling data • Deep reinforcement learning for automated trading and oil price prediction • RNNs for real-time failure detection and forecasting to support immediate decision-making • Deep neural networks for forecasting oil demand, optimizing pricing & production strategies • LLMs to build a private ChatGPT integrated with the enterprise knowledge base • GANs for image enhancement, noise attenuation, bandwidth extension, and compressive sensing • 3D encoder-decoder networks for automatic geological interpretation from large-scale seismic data • VAEs for seismic inversion, providing probabilistic insights into reservoir properties

Education

• UT

  PhD, Computational and Applied Mathematics

• Nankai University

  Bachelor, Computational and Applied Mathematics

Skills

• Python
• Unix
• Apache Spark
• Finite Element Analysis
• Semi-Submersible
• Deep Learning
• Grid Generation
• Dynamic Positioning
• Fpso
• R
• Microsoft Azure
• Machine Learning
• Aqwa
• Data Mining
• Parallel Programming & Computing
• Reinforcement Learning
• Coding
• C/C++
• Fortran
• Naval Architecture
• Wamit
• Hadoop
• Star-Ccm+
• Orcaflex
• Mooring
• Computer Vision
• Big Data Analytics
• Seakeeping