I earned NVIDIA's Certificate of Competency for Building RAG Agents with LLMs. This certificate fits directly into the kind of AI systems I want to build: agents that can reason with documents, retrieve useful context, manage conversation state, and produce more grounded answers.

The training focuses on retrieval-augmented generation, LLM pipelines, vector stores, embeddings, document reasoning, dialog management, and evaluation. It also covers practical tooling around LangChain, Gradio, and LangServe.

Skills strengthened:

• Designing RAG systems that retrieve context before answering
• Using embeddings and vector stores for semantic search
• Handling long documents through chunking and refinement
• Building dialog flows that keep useful state
• Evaluating RAG outputs with a more systematic mindset

This was a good step toward building AI applications that are more reliable than simple prompt-only demos. RAG is especially important for portfolio projects where correctness, source grounding, and context matter.

More info on LinkedIn: NVIDIA certificate post

NVIDIA course reference: Building RAG Agents with LLMs