multimodal data, label-efficient learning, synthetic data generation. University of Washington PhD

Startup Operations: Founding EA & Operations at startups backed by Y Combinator, Sequoia, etc.

Formerly Apple: Apple Intelligence Foundation Models, multimodal research, visual grounding, computer-use agents, UI understanding; formerly Google Research: generative UIs. Berkeley CS PhD.

Former AWS AI engineering leader. Built infrastructure for foundation model training, customization, and inference at scale. Previously worked on Bedrock, SageMaker and Kubeflow; focused on distributed ML systems, RL, and large-scale AI platforms.

Formerly Google Research: Gemini Data; pretraining. Data Mixture Optimization, Bilevel Optimization, Federated Learning. Google Docs.

Formerly SRI International: multimodal perception and learning, human behavior modeling, temporal causal discovery, and real-time ML; clinical assessment, driver monitoring, and visual skill understanding.

Former Member of Technical Staff at xAI, working on multimodal LLMs and agents. Former Tech Lead at Salesforce Research, focused on multimodal LLMs and computer-use agents. UCSC PhD.

Formerly ByteDance/TikTok: Strategy & GTM Lead for global markets, spanning cross-regional expansion (US, EU, UK, APAC), strategic partnerships, and venture investments. Harvard MBA.

Formerly Salesforce Research: Tech Lead for multimodal research, spanning multimodal foundation models, multimodal LLMs for understanding, and unified multimodal modeling.

Formerly Google DeepMind: Gemini Data and Agents; Google's first public agent, Project Mariner. Chemical formulation stability prototype model. Geospatial computer vision. Superfund research: subsurface groundwater contamination modeling

Formerly business operations/strategic finance at EvenUp & Abnormal AI. Technology investor at Tidemark and Warburg Pincus. UC Berkeley Computer Science/Business.

Formerly Google DeepMind: Gemini post-training, robust reward self-improvement, prompt and data optimization, co-creator of OSS Vizier and Optformer. Berkeley PhD.

Formerly Google DeepMind: core architect for Gemini; pioneered the inference and distillation strategies that define the state-of-the-art. CMU PhD.

Founding Recruiter & Head of Talent Acquisition. Formerly Recruiting Researcher at Netflix: AI/ML/DS hiring. Previously managed business development, marketing and product development at LG Electronics. Ewha Womans University Business BA.

Formerly Preferred Networks: over a decade at the forefront of deep learning, spanning neural architectures and learning objectives for generative modeling, unsupervised learning, and representation learning. Introduced Spectral Normalization and Virtual Adversarial Training. Previously Google Brain.

Formerly Noble Machines: Co-Founder and Head of AI; end-to-end autonomy for industrial humanoids, sim-to-real RL, vision-language-action policies. Formerly Apple Special Projects Group: multimodal perception for autonomous systems. CMU Robotics

Prior Meta engineer under AI infra and wearables agentic AI. Focusing on infra readiness and agentic data flywheel for post training. Rice alumni.

Formerly Apple: Agentic RAG architecture for Apple Intelligence, LLM post-training; formerly Google Research: multimodal Neural Architecture Search. SUNY Buffalo CS PhD.