Build and optimize state-of-the-art, resource-efficient deep learning models and multimodal LLM features for deployment on low-power embedded devices. Improve training algorithms, data pipelines, model architectures, and training infrastructure (cloud and edge), and ship models to production at scale using Kubernetes, GCP, and auto-labeling workflows.

Design, train, and deploy state-of-the-art, highly efficient computer vision and multimodal LLM models for low-power edge devices. Improve training algorithms, data pipelines, model architectures, and deployment tooling to ship AI features to millions of cameras. Prototype demos and optimize cloud and on-device inference.