We're at the cutting edge of leveraging AI technologies in energy storage solutions, particularly for enhancing lithium-metal batteries at SES AI. Combining graph neural networks, representation learning, reinforcement learning, generative models, and large language models (LLMs) can offer a comprehensive and powerful approach to not only understanding complex battery chemistry and performance factors but also to innovating in the space of lithium-metal battery technology.SES is a global leader in development and production of high-performance Li-Metal rechargeable batteries for electric vehicles (EVs) and other applications. Founded in 2012, SES is an integrated Li-Metal battery manufacturer with strong capabilities in material, cell, module, AI-powered safety algorithms and recycling. Formerly known as SolidEnergy Systems, SES is headquartered in Boston and has operations in Singapore, Shanghai, and Seoul.

Head of Artificial Intelligence-MLP, Senior Principal AI Scientist, with the joint appointment in the Department of Cancer Immunology, at Genentech Inc.Led a highly experienced AI/machine learning team at Genentech with strong biotech industry experience developing and delivering cutting-edge machine learning/deep learning based models for ten major projects spanning from target discovery, drug development to translational medicine and clinical trials. Two deliverable efforts of mine have been acknowledged and recognized by the Roche Board. Led the pioneering effort and AI/bioinformatics teams to develop and introduce the first AI algorithm in clinical trials at Genentech aimed at addressing the critical research issue of selecting the optimal peptide for personalized cancer vaccine and T cell based therapy for patients. Deliver spatial biomarker discovery from HCC patients sample by AI/ML modeling on spatial transcriptomics data, integrating multiple teams involving biomarker, research, bioinformatics experts. Deliver multimodality AI/ML modeling and spatial modeling on clinical trials patients data to help stratify patients and conduct reverse translational medicine studies to understand tumor microenvironment for biomarker/target discovery in Cancer Immunotherapy. Led the Natural Language Processing team and clinical protocol authoring team to deliver AI protocol Assistant products for assisting internal/external physicians to write and interpret clinical trials protocols. Developed the LLM/vision language model for oncology and invented new methods for knowledge graph to co-optimize with experimental data with the goal of nurturing machines capable of thinking like actual scientists.

Neuroscience-inspired AI on incremental and lifelong learning, preventing catastrophic forgetting problem. Inspired by the hippocampus and EC in the brain, we developed centroids-based memory neural networks for NLP studies.

Working on models of cortex-inspired hierarchical temporal memory

Deep learning NLP research with end-to-end coattention model on reading comprehension

Rotation research in Dr. Loren Looger's lab