Pass the NVIDIA Agentic AI Professional Cert in 6 Weeks

About this course

This course contains the use of artificial intelligence.

BEFORE PURCHASING THIS COURSE:This course uses an AI-generated voiceover for clarity and consistency. Please consider this before purchasing if you prefer human narration.

This is an unofficial, independent course. It is not affiliated with or endorsed by the exam provider.

Preparing for the NVIDIA Agentic AI Professional (NCP-AAI) certification exam? This course teaches the exam's topics one short lesson at a time, every lesson comes with detailed study notes, and every module ends with a quiz of scenario questions in which every answer option is explained.

What's inside

Module 0, Start here: what the exam covers and how the course maps to it, how to choose between the Associate and Professional certifications, and setting up your lab environment for free (a hosted model API or a local open model) or on AWS with a GPU, auto-stop and teardown.

Module 1, Agent Architecture and Design: agents and workflows, the human-agent interface, ReAct, agent-to-agent communication with A2A and MCP, multi-agent patterns, knowledge graphs, and designing for scale and change.

Module 2, Agent Development: prompt engineering and prompt chains, tools and function calling, MCP servers, multimodal inputs, error handling (retries, timeouts, circuit breakers, fallbacks), streaming, and evaluating how an agent decides.

Module 3, Cognition, Planning and Memory: chain-of-thought and task decomposition, plan-and-execute, ReWOO and Tree of Thoughts, short-term and long-term memory, checkpoints and resuming, and learning from feedback with reflection and self-critique.

Read the full description on Udemy on Udemy (opens in a new tab)

Prices, discounts and availability are set by Udemy. We may earn a commission when you purchase through links on this site.