Understand Integration Architecture
Learning Objectives
After completing this unit, you’ll be able to:
- Describe the two-layer integration model in Agentforce IT Service.
- Identify the difference between employee channel integrations and MuleSoft for Flow connectors.
- Explain why integration flexibility is a key differentiator for Agentforce IT Service.
Before You Start
Before you start this badge, consider completing this recommended content.
The Connected IT Challenge
Imagine this: Your employees are working across Slack, Microsoft Teams, web portals, and legacy ticketing systems every single day. Meanwhile, your IT team is managing a complex patchwork of tools, including identity providers, monitoring platforms, project trackers, and competing IT service management (ITSM) systems.
Now think about how you might incorporate AI across all these tools. Without a unified integration layer, your AI agents can only go so far. Sure, they can answer a user’s questions, but they can’t take action. They can’t reset an Okta password, create an epic in Aha, or grant access to SharePoint. The challenge you face is figuring out how to connect everything without having to write and maintain custom code for every single system.
Meet Employees Where They Work
Before you dive into the underlying architecture, it helps to understand the surface level. Agentforce IT Service lets you bring your IT service desk to your employees through three primary collaboration channels.
- Slack: Embeds key workflows to deliver instant, guided self-service. Employees can effortlessly report issues or request services without leaving their workflow.
- Microsoft Teams: An AI-powered collaborative workspace where employees can request services, report issues, and receive real-time notifications.
- Employee Services Portal: A self-service hub for employees to find knowledge articles, track IT requests, and escalate to service reps when AI isn’t enough.
The Two-Layer Integration Model
Agentforce IT Service solves this connected IT challenge using a distinct, two-layer integration model. Let’s go through this real quick.
Layer | What It Connects | Purpose | Examples |
|---|---|---|---|
Layer 1: Employee Channel Integrations | Connects employees to AI agents | Where employees reach out to interact with AI agents | Slack, Microsoft Teams, Employee Services Portal |
Layer 2: MuleSoft for Flow Connectors | Connects AI agents to external backend systems | Where AI agents take action to fulfill requests | Okta, Aha, SharePoint, Miro, Asana |
MuleSoft for Flow
So, how does that second layer work? Enter MuleSoft for Flow.
MuleSoft for Flow gives Agentforce IT Service access to over 200 prebuilt connectors. You can invoke these connectors directly from Salesforce Flow. This lets your AI agents provision accounts, reset access, create tickets, and perform other critical tasks in external systems—all without needing any custom development. Custom code? Doable, but no thanks! Using existing, proven connectors is simpler.
What’s Next?
Now that you understand the foundation of the two-layer architecture, it’s time to explore it in practice. In the next unit, you review the employee channels that your workforce uses to reach agents. Then, you examine the connectors that your agents use to take action in those critical external systems.
