# Executive Demo Guide

## Duration

30 to 45 minutes.

## Main Message

Oracle Deep Data Security protects data at the source. Even when an AI agent, vibe-coded application, BI tool, or dynamic SQL attempts to query more data than it should, the database enforces authorization by end user, role, and context.

## Demo Flow

1. Show the problem: an application or agent uses a powerful connection.
2. Run a risky request: "list all salaries and documents."
3. Show the excessive result in the baseline, when applicable.
4. Enable or explain the data grants.
5. Run the same query as a limited user.
6. Show that unauthorized rows and columns are filtered or masked.
7. Show a manager or HR user with broader authorized visibility.
8. Close with audit and governance evidence.

## Recommended Customer Track

1. `05-legacy-app-ai-extension`: safe legacy modernization with AI.
2. `06-rag-vector-classified-docs`: authorized chunk retrieval before LLM context injection.
3. `07-audit-evidence-data-safe`: evidence, auditing, and governance with Data Safe.

## Value Statements

- "Authorization follows the user, not only the application."
- "Controls live in the database, where the data resides."
- "Applications, agents, and BI tools respect the same data boundary."
- "The policy is declarative, versionable, and auditable."