Richard Wang
Co-founder @ Clad Labs
About
Quit my job. Fun AI projects now.
Work Experience
Aug 2025 - Present (1y 3m)
Building cladlabs.ai
Mar 2025 - Aug 2025 (5m)
Acquired by Tokyo Electron (TEL). - Continuing work on high-dimensional time series embedding and physics-informed AI - Patent filing: Invention Ref: 260260 (63/817,227) - Conference Presentation: AI Agents in few-shot learning for autonomous fault detection and prediction in semiconductor fabrication (Semi-Con, Milpitas 2025)
Aug 2023 - Mar 2025 (1y 7m)
Machine learning research for semiconductor development. • Part of small-team startup; contributed to SoTA technology that led to acquisition by TEL. • Developed full ML data pipelines, training, and inference pipelines, servers, and API from scratch using Docker, Kubernetes, MLflow, Tensorflow, Tensorflow Serving, and Flask. • Led research in generative models for high-dimensional, sparse time-series data from semiconductor fabrication; predictive maintenance and anomaly detection for on-device real time inference. • Implemented novel physics embedding models and used in LLM fine-tuning with text-prompting for natural language description of complex physical processes. • Developed novel mathematical learning for modeling chaotic dynamical systems with contrastive learning and spectral analysis • Research and development into Dynamic Mode Decomposition (DMD), Complex Valued Neural Networks (CVNN), Wavelet analysis, VAEs, transformers, and other SoA deep learning. • Deployed SoTA models for large-scale user base. Production software for internal and consumer use.
Jan 2023 - Apr 2023 (3m)
Snowflake workload optimization. • Contributor at seed-stage startup backed by Greylock, Foundation Capital, and etc. • Automated query optimization on Snowflake, ML for time series modeling on query metadata.
