Introduction to Agent Factory
Oracle AI Database Private Agent Factory (Agent Factory) is a no-code platform that allows both business users and engineers to build, test, and deploy intelligent agents. The platform connects to enterprise data, LLMs, APIs, databases, and tools so teams can create governed assistants and workflows using:
- Pre-built agents
- Custom-built agents
- End-to-end workflows
The application ships with two pre-built agents - Knowledge Agent and Data Analysis Agent, along with a curated collection of pre-built agentic templates to help you get started quickly with intelligent automation.
Before you install Agent Factory for a team, review Architecture and Security, Prerequisites and Installation Modes, and Database Preparation and Grants. These pages explain the deployment components, network paths, database users, secrets, TLS, SSO, and outbound URL controls that production administrators need to plan.
Who Uses These Docs?
| Role | Start with |
|---|---|
| Evaluator | Download Kit, Prerequisites and Installation Modes, and the Quickstart flow for your operating system. |
| Platform administrator | Architecture and Security, Set Up on Linux, Configure Proxy, Collect Diagnostics, and Upgrade Agent Factory. |
| DBA | Database Preparation and Grants, Configure Select AI Framework, and Database Data Source. |
| Security reviewer | Architecture and Security, Manage Users, Configure Single Sign-On, Configure Proxy, and Manage SSL Certificates. |
| Agent builder | About Agent Builder, Agent Builder Nodes, Sample Workflows, and Export and Import Workflows. |
Key Features
No‑code Agent Builder and AI Workflow Automation:
Now anyone can rapidly design, prototype, and deploy powerful AI agents, no engineering background required. With its intuitive drag-and-drop interface and visual abstraction, Agent Builder enables seamless integration of AI and automation, easy connectivity with enterprise systems, and quick custom agent creation using reusable templates.
Whether you need simple workflows, enhanced agents leveraging custom and built-in tools, advanced custom-component agents built-in Python, or complex multi-agent orchestrations, Agent Builder provides the flexibility and capability to meet diverse business needs, making AI accessible and actionable for all.
Knowledge Based AI Agents:
Connect Google Drive, SharePoint, Amazon S3, websites, and file systems, then ground answers with Oracle AI Vector Search for accurate, traceable responses from approved sources.
Data Analysis Agents:
Seamlessly interact with your enterprise databases, intelligently understand table / view structures, and automatically extract valuable semantic insights using advanced LLM capabilities.
The Data Analysis Agent offers features such as context-aware question generation through variation analysis, LLM-powered explanations for complex data, and automatic visualization generation to make insights easy to understand and actionable. With out-of-the-box integration for Oracle Database 19c and later, you can connect instantly and unlock deeper, faster data-driven decision making across your organization.
Models and Interoperability:
Choose embedding and generative models from any of these providers (OCI GenAI, Ollama, OpenAI, vLLM, or Google) and import/export with Agent Spec to reuse your work in LangGraph, AutoGen, or CrewAI.
Security and Governance:
Protect access with SSO and role management, enforce prompt guardrails, and ensure grounded, source‑linked responses with built‑in evaluation before production.