How Oracle’s AI Fusion Application is Constantly Evolving

Oracle Unveils Major Updates to AI Agent Studio and Launches New AI Agent Marketplace

Making enterprise AI easier, faster, and more impactful than ever

At Oracle AI World 2025, Oracle announced a major expansion of the Oracle AI Agent Studio and introduced the new Oracle Fusion Applications AI Agent Marketplace—a big leap toward making enterprise AI easier to adopt and far more scalable.

If you’ve ever tried to bring AI into your business processes, you know the challenges:

·       Too many tools

·       Hard-to-integrate models

·       Skills gaps

·       Security concerns

·       Slow time-to-value

Oracle’s latest updates aim to eliminate those barriers by giving organizations a unified place to build, manage, test, deploy, and now shop for enterprise-ready AI agents that work across Oracle Fusion Cloud Applications.

In simpler terms:

Oracle is turning Fusion Applications into an “AI-ready platform,” where companies can instantly plug in trusted AI agents—whether built by Oracle, partners, or their own teams—to automate work, improve accuracy, and speed up decisions.

Let’s break down what’s new and why it matters.

1.Introducing the Oracle AI Agent Marketplace

Partner-built AI agents you can deploy with a click

Oracle now offers a marketplace inside Fusion Applications, where customers can browse and install Oracle-validated AI agents built by top global systems integrators and software vendors.

Think of it like the App Store—but specifically for enterprise AI agents that automate real business processes.

What these agents can do

Examples from launch partners:

·       Faster sales order processing
Automatically align addresses across CX and ERP systems for cleaner, error-free orders (Apex IT).

·       Automated sales order entry
Let employees create and validate sales orders with natural language instead of manual entry (IBM).

·       Hire-to-retire HR automation
Fetch, update, and synchronize employee data while respecting role-based access control (Infosys).

·       Procurement intelligence
Instantly analyze historical pricing trends to improve supplier negotiations (KPMG).

·       Document content extraction from Box
Pull structured information from Box files directly into Oracle workflows (Box).

·       Invoice collection automation
Improve collection rates using Stripe’s AI-powered payment optimization (Infosys + Stripe).

·       Candidate profile enrichment
Standardize and update resumes using publicly available information (RChilli).

This marketplace allows customers to deploy vetted, secure, production-ready agents—without coding or complex integrations.

Why it matters:

·       10x faster AI adoption

·       Lower onboarding cost

·       Built-in trust and security

·       Domain expertise “baked into” each agent

·       Seamless integration with existing workflows

2. AI Agent Studio Gets Bigger: More LLMs, More Tools, More Control

Oracle AI Agent Studio is the platform developers and business teams use to build and deploy their own AI agents.

This update makes the studio far more powerful and flexible.

Support for all major LLMs

You can now choose from:

·       OpenAI

·       Anthropic

·       Cohere

·       Google

·       Meta

·       xAI

This means customers can pick the best model for cost, speed, or accuracy.

3. New Integration & Extensibility Features

Model Context Protocol (MCP)

MCP allows agents to safely connect to external systems and tools—unlocking data outside Oracle without custom glue code.

A2A (Agent-to-Agent) Cards

Agents can now talk to each other and pass context, enabling multi-agent collaboration.

Secure Credential Store

API keys and authentication tokens are securely stored so agents can access external data without exposing credentials.

4. New Observability & Evaluation Tools

Companies need AI agents they can trust. Oracle added enterprise-grade monitoring:

·       Real-time dashboard (latency, sessions, error rates)

·       Tracing for debugging agent workflows

·       Token usage reporting for cost control

·       Quality and correctness scoring

·       Safety evaluations

This turns AI into a measurable, governable enterprise asset—not a black box.

 5. Better Prompt & Knowledge Management

Oracle introduced:

·       Prompt libraries

·       Topic management

·       Version control for agent prompts

This keeps AI consistent across teams and reduces risk from prompt fragmentation.

6. More Agent Templates and AI Builder Assistants

You can now start from pre-built template agents or use a guided assistant that helps you generate:

·       Topics

·       Prompts

·       Tools

·       Workflows

This dramatically lowers the skill barrier for agent creation.

7. Multimodal and RAG Enhancements

Agents can now reason over:

·       Documents

·       Images

·       Tables

·       External content stores like SharePoint

This opens the door to more sophisticated analysis (e.g., reading invoices, reviewing designs, analyzing spreadsheets).

8. Advanced Workflow Automation

New workflow capabilities include:

·       Deterministic agent workflows (predictable, rules-based outcomes)

·       Chained workflows for multi-step processes

·       Agent nodes for dynamic reasoning

·       Human-in-the-loop approvals

This blends automation and oversight—ideal for finance, HR, legal, and procurement.

9. A Growing Network of 32,000+ Certified AI Experts

Oracle now has a global ecosystem of certified partner experts trained specifically on:

·       Building AI agents

·       Using Agent Studio

·       Deploying agents into Fusion Apps

This makes it easier for customers to find skilled implementation partners who know the tooling and follow Oracle’s security standards.

What This Means for Enterprises

1. Faster AI adoption

Deploy agents instantly instead of starting custom projects from scratch.

2. Lower risk with validated agents

Security, governance, and best practices are baked in.

3. Real productivity gains

AI agents automate tedious tasks across finance, HR, supply chain, customer experience, and more.

4. Customization without complexity

Low-code/no-code tools allow organizations to tailor agent templates to their exact needs.

5. Better performance and lower cost

Teams can pick the most cost-effective LLM and monitor usage in real time.

6. Future-proof foundation

Oracle is building an AI ecosystem where agents, workflows, and partner solutions evolve continuously.

Example Use Cases Across the Enterprise

Finance

·       Automated invoice processing

·       AI-driven payment collections (Stripe integration)

·       Procurement intelligence (KPMG)

·       Error-free sales order entry

Human Resources

·       Automated employee profile updates

·       Candidate profile enrichment

·       Onboarding workflows with human-in-the-loop approvals

Supply Chain & Operations

·       Address alignment and shipping verification

·       Multimodal document comparison for logistics

·       Automated material master updates

Customer Experience

·       AI assistants that create and validate sales orders

·       Cleaner, enriched customer contact data

·       Automations triggered by unstructured content from Box

The Big Picture

Oracle isn't just adding AI to Fusion Applications—it’s transforming Fusion into a full AI platform.

The combination of:

·       AI Agent Studio (build + manage your own AI agents)

·       AI Agent Marketplace (deploy partner-built agents instantly)

·       Enterprise-grade governance + observability

·       Deep Fusion integration

…creates one of the strongest end-to-end AI ecosystems in the enterprise application market.

It’s AI for real business processes—not experiments.

As enterprise AI evolves, it is easy to feel overwhelmed by rapid changes, new capabilities, and shifting best practices. Our team understands these challenges, and we are here to help you navigate them with clarity and confidence. If you are exploring how AI agents can strengthen your Oracle environment or support your broader transformation goals, we would be glad to guide you.

Connect with iArch Solutions to schedule an AI Readiness Assessment and build a clear, low-risk roadmap for bringing intelligent automation into your organization.

Schedule a Discovery Call
Next
Next

Essbase Marketplace Update: What’s New in Stack Release 21.7.3.0.2