✦ Velza Academy · Free Self-Assessment

Where do you really stand on the NVIDIA-Certified Professional: Agentic AI (NCP-AAI) exam?

Most people preparing for this exam study what feels familiar. The exam grades what the official blueprint weights. The free readiness scorecard shows you the difference in about five minutes.

Rate yourself topic by topic across every official exam domain
Your scores are weighted exactly the way the exam is, so a weak heavy domain jumps out first
Charts show your readiness per domain at a glance
Progress saves in your browser, so you re-score as you study and watch the gaps close

Built by Mike Wheeler, O’Reilly author, 500,000+ learners taught

Get your free scorecard

Enter your email and we’ll send your scorecard link. Click the confirmation link to open it.

Free, and yours to keep. You will also get occasional Velza Academy updates about this certification. Unsubscribe anytime, in one click, from any email.

✦ Inside the scorecard

Your readiness, charted as you slide

Slide your honest rating for every topic and the picture updates live: readiness per domain, weighted study priorities, and the chart of where you stand.

The NVIDIA-Certified Professional: Agentic AI readiness scorecard showing the domain chart, weighted study priorities, and topic sliders

The actual NCP-AAI scorecard, mid-assessment.

What the NCP-AAI exam actually is

These details come from NVIDIA's current certification page and linked study guide. Read the official exam page alongside this scorecard so you are working from the source.

What it is: An intermediate-level credential validating the ability to architect, develop, deploy, evaluate, and govern advanced agentic AI solutions.
Who it is for: Software developers and engineers, solution architects, machine learning engineers, data scientists, AI strategists, and AI specialists with production-level agentic AI experience.
Questions: 60–70
Duration and delivery: 120 minutes; online with remote proctoring
Cost and validity: $200; valid for two years from issuance and renewable by retaking the exam
Passing score: NVIDIA does not publish a numeric passing score on the certification page. Use the published domains to prioritize your study.
Study-guide blueprint: Agent Architecture and Design 15%; Agent Development 15%; Evaluation and Tuning 13%; Deployment and Scaling 5%; Cognition, Planning, and Memory 10%; Knowledge Integration, and Data Handling 10%; NVIDIA Platform Implementation 7%; Run, Monitor, and Maintain 7%; Safety, Ethics, and Compliance 5%; Human-AI Interaction and Oversight 5%. The linked PDF totals 92%. NVIDIA's page displays conflicting weights totaling 98%; this scorecard does not invent replacements.
Prerequisites and language: One to two years in AI or ML roles plus hands-on production agentic AI work and the architecture, orchestration, tools, evaluation, observability, deployment, interface, and guardrail experience described by NVIDIA; English.

Mike Wheeler's matching Udemy practice-test course artwork

After your self-assessment

Practice the exact certification next

Use the free scorecard to find your gaps. Then use Mike Wheeler's matching Udemy practice tests to work through exam-style questions and explanations.

See the practice tests

Course artwork shown above is from Mike's Udemy practice-test course. It is not a vendor-issued certification badge.

✦ The blueprint, weighted

The published study-guide domain weights

These are the ten weights in NVIDIA's linked study guide. They total 92%, while the public page displays a conflicting 98% total. The scorecard preserves the PDF values and normalizes their total only for the overall self-rating.

Agent Architecture and Design15%

Agent Development15%

Evaluation and Tuning13%

Cognition, Planning, and Memory10%

Knowledge Integration, and Data Handling10%

NVIDIA Platform Implementation7%

Run, Monitor, and Maintain7%

Deployment and Scaling5%

Safety, Ethics, and Compliance5%

Human-AI Interaction and Oversight5%

✦ How it works

Five minutes to an honest study plan

Rate yourself

Slide through every topic the exam tests. No login, no grading, just your honest read.

See your weighted gaps

Charts show your readiness per domain, weighted the way the exam actually scores.

Study where it pays

Put your next hour on the weakest heavy domain, then re-score and watch the gap close.

✦ Your instructor

Mike Wheeler, founder of Velza Training and Consulting

Built by Mike Wheeler

Mike Wheeler is an AI and Salesforce trainer and O’Reilly author who has taught well over 500,000 global learners on platforms such as edX, LinkedIn Learning, Pearson, and Udemy. He built this scorecard for people aiming at NVIDIA’s Agentic AI exam.

500,000+ global learners taught O’Reilly author Free reusable scorecard