Agentic AI for Oracle AI Database Professional practice questions and study topics

Exam details

Official exam duration

90 minutes

Official question count

50

Official passing score

68%

Practice questions

100 questions across two practice tests

Exam fee

See Oracle exam subscription pricing

Exam topics

• Explain Oracle AI Vector Search Foundations 8%
• Apply Vector Distance and Indexing 32%
• Build Retrieval-Augmented Generation Solutions 24%
• Design Select AI Agent Solutions 16%
• Configure Agentic Interfaces and Tool Access 12%
• Govern Private Agent Factory Deployments 8%

Practice format

Practice includes questions, explanations, and two mock tests. Interactive tasks and provider labs are not simulated.

Published exam question formats: Multiple choice.

Study notes and exam details

• The public English exam page lists proctored online Multiple Choice delivery, 90 minutes, 50 questions and a 68 percent passing score. No dated guide edition or exam fee is displayed.
• Both practice forms use 4/16/12/8/6/4 questions at the published 8/32/24/16/12/8 domain weights. All 14 English objective bullets are mapped with editorial identifiers, not official Oracle numbering.
• The overview does not specify a separate response-type mix. Practice includes multiple-select study exercises with complete-set grading without claiming an official item distribution.
• Vector-search behavior and index support depend on the database release and deployment. Exact baselines should explicitly request EXACT; the current guide notes that ADB-S can attempt approximate search when EXACT is omitted.
• Select AI Agent, managed ADB MCP/A2A interfaces and Private Agent Factory have distinct configuration and capability boundaries. Enabling MCP does not automatically enable A2A. Private Agent Factory data may still be sent to a configured external model or tracing provider.

Official references

• Official certification details