Agentic AI is the key to achieving increased automation, speed, and efficiency, replacing cumbersome manual processes with split-second decision-making and insights powered by autonomous reasoning.
To realize the potential of agentic AI, businesses need fast, flexible ways of building and managing agents. They also need AI governance, rock-solid AI security, and 24/7 visibility into everything agents do.
At Boomi, we’re providing all these capabilities and more with our latest updates to the Boomi Enterprise Platform. In our recent and upcoming product releases, we’re giving business and IT teams unprecedented flexibility and control for integrating data, building and governing agents, accessing tools, and putting the power of agentic AI to work in ways that deliver real business impact.
Simply put, we’re giving customers agentic transformation they can trust.
Here’s a look at what’s new.
Extending Agentic Access to Tools and Systems Securely
Introducing Delegated Authorization
It’s essential that when agents act, they do so with proper authorization, never exceeding the access rights of the users or processes on whose behalf they are taking action. And of course, without authorization, agents can’t perform many actions that would otherwise be useful for their mission. Agents need to be able to act on behalf of users. But how?
Developers could hardcode users’ authentication credentials in agents, but this manual approach is error-prone and ultimately unmanageable at scale. After all, if a user’s credentials changed or if their access privileges evolved in line with their job role, all the agents that user has access to might need to be revised. And someone somehow would need to keep track of which agents have been programmed with which users’ authentication credentials for which resources.
A better approach is to automatically delegate authorization from a user to an agent running on the user’s behalf. By tying agent authorization to centrally managed authentication controls, IT teams retain control over and visibility into the authorized actions of all AI agents. Centralizing control also reduces the risk of shadow IT and agents somehow circumventing official authentication controls in their work.
In our November release, we introduced delegated tool authorization for agents, tying an agent’s access rights for tools directly to a user’s own credentials (managed through an open standard called OAuth). By supporting delegated authorization, the Boomi platform:
- Enables the creation of personal productivity automations, connecting agents securely to user-specific accounts in external services, such as business applications, software services, databases, and more.
- Improves the precision of access controls by relying on user credentials rather than system credentials, which often provide broader, more generalized access rights.
- Improves the precision and accuracy of logs and audit trails, improving the traceability of agent activity.
- Blocks agents from achieving unauthorized access to IT resources.
- Reduces security risks by eliminating the need to store and manage system credentials across tools.
Making It Easy for Agents to Access Popular Tools and Cloud Services
Not only have we made it easier for agents to access tools and other IT resources securely. We’re also accelerating and streamlining agent design and deployment by enabling agents to connect to a new, pre-configured suite of Application Tools. These pre-configured tools make accessing Slack and other applications fast and easy. Simply add the tools to an agent, and they’re instantly available. Combined with delegated authorization, these new tools streamline the creation of personal productivity workflows. And they help democratize agentic transformation, giving individual business users the power to build agentic solutions that streamline daily work.
We’re also making it easier for agents to access a broader range of data with our newly enhanced Data Connector Agent, which now supports Databricks, Amazon S3, Amazon Redshift, and Amazon RDS as data targets, extending the already impressive range of Boomi Data Integration.
Improving the Performance and Reliability of MCP Servers
The open-source Model Context Protocol (MCP) has emerged as an industry standard for providing agents with access to tools ranging from databases to business applications to command-line interfaces. To ensure that critical business processes that rely on MCP never fail, we’ve added MCP multi-node availability, supporting enterprise-grade scale, performance, and reliability for agents and tools.
Cross-platform Cohesive User Experience
Making Boomi Processes Callable Through MCP
The Boomi platform brings everything together: data, agents, APIs, and automation. We want to make it easy for customers to combine these elements in new and creative ways to solve business problems.
To that end, we’ve introduced an MCP Server Connector for Boomi Integration. Inserting this shape at the beginning of a Boomi process makes the process discoverable and executable by AI agents through MCP. That means agents can now run Boomi processes as part of their work, making Boomi integrations and workflows part of the work that agents can manage and automate.
Auto-Importing APIs as Tools in Boomi Agentstudio
Boomi Agent Designer now natively integrates with API Control Plane, Boomi’s solution for API gateways and APIs across your organization. Through this integration, you can now automatically import APIs as tools for use by agents. Agent Designer uses AI to automatically generate input parameters based on API metadata or descriptions, making it quick and easy for agents to call these APIs and take advantage of their capabilities as agentic tools.
Creating API Developer Portals for Managing and Accessing APIs
APIs are foundational to agentic AI. They function as doorways to the data and operations agents need for getting work done.
Many organizations suffer from API sprawl. They have hundreds or even thousands of APIs, but these APIs aren’t centrally managed or catalogued, leaving them out of reach for agentic transformation. To lay the foundation for an MCP catalog of APIs for use by agents, Boomi now allows customers to create multiple customizable API developer portals through Boomi API Management. These Boomi-managed developer portals will help organizations better catalog and manage their APIs and drive API adoption.
One benefit? Even more well-managed data becomes available for agentic transformation.
Supporting Structured I/O for Agents
Boomi customers can now create more predictable agentic automations using reliable data formats with new structured I/O agent support in Boomi Integration.
Boomi Integration’s Agent Step, which enables an integration process to call an agent as part of its workflow, now supports structured JSON profiles for both requests and responses. Supporting these structured data formats eliminates the need for manual scripting and message parsing, making it much easier to build, map, and run agents that handle complex data. This innovation frees developers to develop more traditional application-focused agents without having to rely on conversational workflows.
Expanding Access To SAP Data Through Change Data Capture (CDC)
For many organizations, SAP serves as an invaluable source of data for agents. We want to make sure that customers have a variety of ways of accessing data in SAP, so they can choose the data integration approach that best supports their agentic use cases and need for advanced analytics.
We’ve expanded our SAP Data Connector to support Change Data Capture (CDC), so that SAP data can be replicated to cloud data warehouses for enterprise-wide operations and agentic workflows.
End-to-end Observability and Management
As agent and agentic workflows automate and streamline more work in organizations, agent observability and governance becomes even more important. In upcoming releases, Boomi customers will gain the ability to:
- Monitor agent tools in one place with new agent tool management and analytics capabilities to understand what tools are being used across agents, spot dependencies, and analyze agent interactions.
- Develop a secure, reliable, and transparent AI agent ecosystem using agent session logs, gaining visibility into agent activities and actions to identify failure root causes within each agent.
Log in to the Boomi platform to see some of these features in action or request a personalized demo.