nvidia/skills

Agentic AI

Build and operate agentic systems, including RAG workflows, evaluation harnesses, tool use, policy, sandboxing, agent configuration, and agent workflow automation.

Physical AI

Build and validate AI workflows that interact with the physical world, including robotics, autonomy, simulation, synthetic data, and embodied AI systems.

Vision AI

Build vision and video AI workflows for analytics, search, summarization, alerts, real-time understanding, and model integration.

Conversational AI

Build and deploy speech, voice, dialogue, and multimodal conversational AI workflows, including real-time assistant and voice-agent experiences.

Simulation and Modeling

Create, run, validate, and analyze simulation and modeling workflows for systems, environments, processes, and physical phenomena.

Data Science

Prepare, analyze, explore, and model data for AI and accelerated analytics workflows that are not specific to one training, inference, or agent task.

Training AI

Build, validate, tune, and maintain large-scale model training workflows, including distributed training, model support, training validation, performance tuning, and training-stack CI.

Inference AI

Deploy, serve, optimize, and evaluate models for production inference, including quantization, runtime performance, and serving configuration.

Decision Optimization

Formulate and solve routing, scheduling, and numerical optimization problems using optimization solvers, APIs, services, and deployment workflows.

GPU Development

Develop, tune, profile, and integrate GPU kernels and accelerated code, including CUDA-adjacent workflows, autotuning, low-level performance work, and framework integration.

Quantum Computing

Build, run, and validate quantum and hybrid quantum-classical applications, including CUDA-Q development workflows.

Infrastructure

Deploy, operate, monitor, and troubleshoot clusters, runtimes, services, networking, and platform infrastructure for accelerated and AI workloads.

Other skills