Agentic AI Explained
A short NVIDIA overview of what makes AI "agentic": the perceive-reason-act-learn loop, design patterns, agent roles, and real uses. A conceptual primer, not a coding course.
Go to the course on NVIDIA Deep Learning Institute- Provider
- NVIDIA Deep Learning Institute
- Price
- Free
- Certificate
- Certificate not statedThe course page does not mention a certificate; NVIDIA gives them only for select courses.
- Level
- Beginner
- Time
- 1 hour
- Format
- Self-paced lectures and videos
- Checked
- 2026-11-05
Free. Needs an NVIDIA account.
01
What you will learn.
- 01Agentic AI
- 02Agent architecture
- 03Agentic design patterns
- 04Multi-agent roles
- 05Agent frameworks
More inAI agents courses
02
Before you enroll.
From the course page, checked 2026-11-05.
- You need
- Basic programming (especially Python), maths or data science, or another technical background
- Some knowledge of machine learning, automation, or AI helps
03
Similar courses.
Agents and Workflows
A no-code course on handing structured work to agents in ChatGPT: give context, define outputs, set boundaries, and review results. OpenAI recommends taking AI Foundations and Applied AI Foundations first.
FreeFree badgeIntermediate
Securing Agents with NemoClaw and OpenShell
Build an agent loop from scratch, then add reliable tool calls, multi-agent routing, and sandboxed execution. NVIDIA NemoClaw is the reference stack and OpenShell runs the agents securely. Free and code-heavy.
FreeFree certificateIntermediate
Building RAG Agents with LLMs
Build and evaluate a RAG agent that answers questions about a set of research papers, using LangChain pipelines, embeddings, and vector stores. Paid ($90), with an assessment and certificate at the end. It assumes intermediate Python and some deep learning.
PaidPaid certificateIntermediate