Machine Learning Jobs in Sunnyvale, CA

The Staff Machine Learning R&D Engineer will innovate machine learning algorithms for spatial computing, collaborate with teams, and translate research into production solutions.

Lead research and production deployment of ML and computer vision systems for 3D capture and reconstruction. Improve camera pose estimation, mesh/surface reconstruction, texturing, semantic understanding, and real-time capture guidance. Apply deep learning, multi-view geometry, and neural rendering techniques; collaborate across hardware, software, and product teams; and manage the full ML lifecycle from data collection to production monitoring. Write high-quality Python and C++ code.

Develop and optimize computer vision models for drones, improve ML pipelines, collaborate on R&D, and tackle advanced technical challenges.

Design and implement distributed cloud GPU training for ML, build end-to-end pipelines, collaborate on data problems, and uphold ML best practices.

Develop and apply deep learning and data analysis techniques to create scalable vehicle simulation models, collaborating with teams to advance autonomous technology.

The intern will support the data science team by gathering data, performing exploratory analysis, building machine learning models, and developing visualizations to aid business insights.

Develop and lead evaluation systems, benchmarks, and large-scale data pipelines for Waymo foundation models and embodied AI agents. Evaluate models across pretraining, supervised fine-tuning, and reinforcement learning for quality, safety, and realism. Collaborate across teams to bring innovative machine learning technologies into production. The role requires strong ML engineering, deep learning, distributed systems, programming, analytical, and debugging skills.

Build, train, fine-tune, and evaluate computer vision and scene-understanding models for offline autonomous-driving measurement. Improve accuracy and generalization across vehicles, markets, and conditions using large models, temporal context, and multi-task learning. Define ground truth, benchmarks, metrics, and statistically defensible evidence for validation and safety cases. Collaborate across modeling, evaluation, data, and simulation teams while mentoring engineers and driving significant modeling workstreams.

Lead data-centric work for embodied AI: mine and curate fleet driving data, build large-scale enrichment and labeling pipelines, train and fine-tune foundation models, and design rigorous offline and closed-loop evaluation to improve safety and generalization across cities, vehicles, and sensor rigs.

Build, train, and fine-tune offline scene-understanding models adapted from on-vehicle and foundation models. Improve accuracy and generalisation across vehicles, geographies, and conditions. Exploit offline compute and temporal context, benchmark models with statistically defensible metrics for validation and safety, and lead cross-functional work while mentoring teammates.

Design and implement machine learning algorithms for robotic manipulation, focusing on RL and IL. Collaborate across disciplines to ensure integration and successful deployment.

Research and model performance of Lightmatter-enabled hardware systems for ML training and inference. Build performance modeling tools, design rack-scale reference platforms, collaborate with product and engineering, publish academic and industry research, and act as the in-house ML systems expert.

Lead RTL development and microarchitecture for an AI-enabled RISC-V CPU core. Responsibilities include designing vector and matrix-processing engines, optimizing power, performance, area, and timing, developing detailed specifications, supporting functional verification, debugging simulations, and partnering on physical design validation. The role requires expertise in high-performance processor architecture, datapath design, vector and matrix processing, and hardware design languages.

Develop high-performance software for silicon test pattern generation, silicon application, failure analysis, and yield learning. Create parallel graph traversal and analysis methods, collaborate across technical teams, and apply LLMs, RAG, graph-based machine learning, and reinforcement learning to electronic design automation solutions. The role also involves large-scale model training, data processing, agentic applications, and advanced multi-agent and vector database systems.

Develop end-to-end perception solutions for autonomous driving, leveraging deep learning and data-driven approaches to enhance system accuracy and efficiency.

Lead development of 3D geometric vision and world foundation models for autonomous driving. Responsibilities include designing and training deep-learning models, building scalable sensor-data, auto-labeling, SLAM, reconstruction, evaluation, and ground-truth pipelines, and deploying models on production vehicles. The role requires technical direction across geometric vision, multimodal sensing, distributed training, neural rendering, 3D representations, and real-time inference.

Build and productionize large-scale ML systems for search, ranking, and personalization. Handle petabyte-scale data pipelines, design transformer/embedding-based retrieval and real-time model serving, track metrics, collaborate cross-functionally, and mentor engineers to improve relevance and customer discovery.

Lead an ML Platform engineering team building large-scale GPU cluster architecture, training and inference orchestration, and performance optimization. Own design, scaling, scheduling, networking, and resource management while partnering with research and stack teams to accelerate production deployment and drive hiring and mentorship.

The role involves building and managing data infrastructure and pipelines for machine learning, automating model retuning, and ensuring efficient data flow from collection to model training.

Build and deploy ML systems for search, recommendation, ranking, retrieval, matching, and personalization. Work on query understanding, embeddings, NLP/LLMs, end-to-end ML pipelines, experimentation, and monitoring to improve relevance, engagement, conversion, and marketplace efficiency in a consumer local-search product.

Develop and deploy state-of-the-art ML models for predicting interactive agent behavior and building controllable agents for closed-loop simulation. Conduct research in generative sequence modeling, transformers, diffusion models, and sequential decision-making, and collaborate with planning and autonomy teams to produce deployable solutions for real vehicles.

Lead end-to-end ML model development for safety-critical driving behaviors, design experiments, improve data quality, deploy models and metrics fleet-wide, and communicate results across product and engineering teams.

Build, train, and deploy production ML models for the electric grid. Collaborate with ML engineers, data scientists, and software engineers across multimodal, NLP, retrieval, and agentic AI domains. Operationalize model training and serving at enterprise scale and stay current with ML research.

Design and deliver production-quality Metropolis AI solutions using Cosmos foundation models. Identify model limitations, generate synthetic data, refine tuning and architectures, partner with Cosmos and cross-functional teams, open-source research artifacts, and advance generative and simulation-based perception for physical AI deployments.

Research and build large-scale neural reconstruction and generative models to complete and synthesize photorealistic 3D spaces from sparse/noisy captures. Design and optimize training and inference pipelines for multi-view, point cloud, mesh, and video data, translate prototypes into production systems, define evaluation benchmarks, and collaborate with product and engineering teams to integrate models into Matterport's digital-twin platform. Contribute to publications, patents, and the strategic roadmap.