Boomi Delivers Agentstudio for AI Agent Development
DALLAS — Joining the parade of recent tech conferences, the major topic of the sessions and conversations at Boomi World 2025 was all about AI agents.
Boomi announced several interesting new capabilities, but by far the major topic at the event here was the new Agentstudio component for developing and governing AI agents.
AI Agents Center the New Boomi Platform Architecture
AI agents work with other components of the Boomi platform, such as integration, orchestration and data management, to reduce manual effort and improve workflow automation.
The new Boomi platform architecture positions AI agents in a central role, leveraging platform components for integration and automation, data management and API management.
Superhumans
Introduced by a robot named B1B, Boomi chairman and CEO Steve Lucas’s keynote opened the conference. Lucas, who recently published a book on the impact of generative AI (GenAI), encouraged customers to initiate AI projects if they haven’t already done so.
Lucas said that AI is not going to replace humans. Instead, it’s going to turn us into “superhumans” by helping us do our jobs better.
Lucas said that 33,000 agents are already in production across the Boomi customer landscape since they initiated their AI program about a year ago. The most popular agent is Scribe, which automatically generates application documentation.
Boomi has also included multiple agents within the product to assist application developers, he added.
Integration, orchestration and automation remain foundational capabilities of Boomi, Lucas said. But GenAI introduces a transformational level of innovation that significantly changes the technology landscape. He urged customers to invest in AI to remain competitive.
Agents, in particular, will change how we interact with our IT systems, he said. We will not be logging into our various applications to check for updates or new information. The agents will do that for us.
Lucas sees a divide coming between “the AI haves and the AI have-nots,” and said the sooner an organization gets started with AI, the better they will be positioned in the market.
Boomi Chief Product & Technology Officer Ed Macosky told The New Stack how he answers questions from customers about how they should get started with AI.
“Customers are somewhat uncertain about AI and focus on the tools and the technology, and try to understand it all from a bottom-up approach. I tell them to first focus on the opportunity that you believe AI can help you with and work backwards from there to the technology.”
Boomi Agentstudio
Agentstudio, the newest capability of the Boomi platform, creates, manages and governs AI agents, including third-party agents.
Agentstudio creates agents and automatically registers them with the Agent Control Tower, which provides observability and governance for them.
Third-party agents, such as those developed using AWS Bedrock, can also be registered with the Control Tower for consolidated discovery, monitoring and policy enforcement capabilities.
“We want to be open and composable, and govern agents developed using other frameworks. We also want to help you bridge all the emerging standards in the agent space, such as MCP (Model Context Protocol) and agent-agent protocol,” Macosky told The New Stack.
Boomi will contribute to the development of all standard protocols in the AI agent ecosystem through its membership in AGNTCY, he added. AGNTCY is an open source collective for interagent collaboration.
Mani Gill, vice president of product at Boomi, walked conference attendees through a demo of Agentstudio, creating and deploying a new agent in a supply chain workflow that automatically finds a new shipping service when the default shipping service isn’t available.
An AI Dynamic Pricing Agent in a Boomi workflow.
AI agents included in a Boomi workflow automate manual tasks such as searching through and analyzing various sources of data to dynamically take action, like calculating new price points or finding a new shipping supplier.
Agentstudio uses the same drag-and-drop, low-code API as the Boomi integration component to develop an agent and drop the agent into a Boomi workflow.
AI Partnership With AWS
Rahul Pathak, vice president, data & AI GTM at AWS, joined Lucas on stage to announce a partnership between Boomi and AWS for AI agents.
Boomi customers can develop an AI agent using AWS Bedrock, Pathak said, and import it into the Agentstudio Control Tower for monitoring and governance. Boomi applications can then discover the agent and integrate it into a workflow.
The Boomi/AWS platform includes integration with Q Business and Q Developer for content integration and developer support, Sagemaker for building and training models, and Boomi Enterprise for integration and orchestration.
Boomi/AWS AI Agent Platform.
The AI Agent Developer Experience
Over the years, Boomi has evolved its developer experience from low code only to low code and pro code, Macosky told The New Stack.
“We’ve been on this journey in which we focus a lot on the low-code developer,” he added. “And we’ve learned through our journey that it’s important to offer a great experience to both low-code and pro-code developers.
“For GenAI, pro-code developers can plug their tools into the Boomi environment through the agent designer,” he said.
“You can also build agents in Bedrock, plug them into the Boomi Control Tower environment, and then build around them using the low-code interface,” Macosky added.
Boomi is offering this comprehensive developer experience for Agentstudio as well, he said.
API Management
Boomi also announced new capabilities for API management, including discovery, security and design features.
Ann Maya, Boomi’s EMEA CTO, told The New Stack that she’s most excited about the new API capability for federated API discovery and management.
“We’re seeing a lot of excitement on the API control side,” she said. “The API control plane is laying the foundation for what we will need to worry about tomorrow in the agentic space as well.
“If you look at the type of breaches that are happening almost every day, the number one attack vector for APIs is shadow and zombie APIs,” Maya said.
That’s why it’s so important to have a federated API management capability, she said, so you can discover all the APIs that need protection across the enterprise, regardless of which gateway or combination of gateways you are using.
Boomi API Gateway for MCP.
The API attack vector will increase with the APIs for agentic AI, she said, and that’s why Boomi has announced an API gateway specifically for MCP.
Data Management
New capabilities in the data management part of the Boomi platform also focus on delivering high-quality data for training large language models (LLMs) and for AI agents to consume.
In his keynote, Lucas said that the quality of GenAI results is fundamentally determined by the quality of the data the AI applications consume.
Boomi is adding contextual data via MCP servers and simplifying data pipelines for retrieval-augmented generation (RAG) processes that enhance the output of LLMs, Macosky said.
Such data management capabilities enable AI agents to interact with data integrations and data sources, improving data quality.
Boomi is unifying its data hub and data integration capabilities into a single platform, Macosky said, to provide a broad set of data connectors and a semantic memory context for agents.
“Customers are often concerned that their data is bad, but if you work from the perspective of a business opportunity, using a very discreet, explicit use case that will provide value, you can start by applying the data management technology to that use case, train the LLM, and develop an AI agent for it,” Macosky told The New Stack.
Integration and Automation
Although conference attendees reported using different parts of the Boomi platform, such as data management or API management, Macosky said the majority use the platform for its original iPass capabilities.
As seen in the updated Boomi platform architecture, Agentstudio leverages the foundational capabilities of the Boomi integration and automation components.
And similar to the enhanced capabilities in the API management and data management components, improvements and enhancements to integration and automation are strongly aligned to support AI agents.
Boomi has added co-pilots for designing, documenting and troubleshooting workflows, revamped the low-code developer experience and added a set of AI connectors.
In his keynote opening the second day of the event, Macosky announced an additional deployment option of a dedicated cloud environment, in addition to the public cloud and managed cloud service options. Boomi also now supports an embedded option for partners, he said.
All deployment options include cost-effective auto-scaling and a 97-point security rating from Security Scorecard.
Macosky also announced that Boomi will be expanding its cloud footprint to Europe and the Middle East.
