Example Projects for Agentic AI
These example projects demonstrate how to build, evaluate, and deploy AI agents using NVIDIA AI Workbench. The projects showcase:
- Agent architectures that connect large language models to tools and external data.
- Agentic Retrieval Augmented Generation (RAG) for autonomous information retrieval.
- Evaluation techniques for measuring agent quality, including LLM-as-judge approaches.
- Model customization with parameter-efficient fine-tuning and reinforcement learning.
- Safety and security hardening for agents before production deployment.
These examples are ideal for developers and organizations looking to:
- Learn agent development through hands-on, notebook-based modules.
- Build agents with NVIDIA NIM microservices, Nemotron models, and the NeMo platform.
- Evaluate, customize, and secure agentic systems for production use.
The Build An Agent Workshop is also available as an NVIDIA Brev Launchable for one-click deployment on a cloud GPU.