From application integration to agent orchestration

Boomi started as a tool for connecting enterprise systems — typically Salesforce with Workday, SAP with Oracle. Over 15 years, it has accumulated hundreds of millions of integration patterns and serves over 30,000 customers worldwide. But now, according to CEO Steve Lucas, the market is changing radically.

"Two years ago we talked about the application economy. Today we talk about the agent economy," Lucas said in an interview with Computer Weekly. There are now over half a million AI models on the Hugging Face platform, and according to Lucas, there could be tens of millions within two years — specialized by industry, company, or even tailored to individuals.

The key question for businesses is: How do you deploy all these agents, adapt them to your own data, and most importantly — keep them under control? And this is exactly where Boomi wants to play a crucial role.

What Boomi actually offers in the AI agent space

Over the past two years, Boomi has significantly shifted from "mere integration" to full-fledged AI agent management. Today, the portfolio includes several key products:

Agentstudio — a central interface for creating, managing, and orchestrating all AI agents in the company. It allows you to define what each agent can do, what data it can use, and how it makes decisions.

Platform Agents — pre-built AI agents for specific tasks: integration design, personal data detection (PII), process automation. The integration design agent, for example, can create connections between systems based on a natural language description — just write "create employee onboarding" and the platform knows which systems are typically used and how to connect them.

Boomi Connect — a set of over 1,000 connectors that give agents controlled access to enterprise applications. Notably, it supports the MCP (Model Context Protocol), which standardizes how AI agents communicate with external tools and data.

Orchestrate — a visual tool for building agent workflows, where individual agents collaborate on more complex tasks — from order processing to loan approval.

Why data matters more than the model

In the interview, Lucas repeatedly emphasized one rule that holds true regardless of how advanced models become: garbage in, garbage out. "If you don't train the model on the right data and fine-tune it properly, it doesn't matter how sophisticated it is," he says.

Boomi has therefore bet on its own large language model trained on hundreds of millions of anonymized integration patterns collected by the platform over a decade of operation. The result is a system that understands how companies actually connect their applications — not in theory, but based on what thousands of organizations have actually done.

For mid-sized companies (with revenues of $500 million to $1 billion), this is a key advantage. They don't have an army of ML engineers or data lakes like Amazon or Walmart. Boomi wants to give them a tool where they select a data source, pick a model, and click "train."

What the numbers say: Boomi in 2026

Boomi now reports over 30,000 customers globally. According to the Gartner Magic Quadrant for iPaaS 2026, Boomi is repeatedly rated as a Leader, confirmed by awards from ISG and Nucleus Research, which measured a 97% return on investment with payback within 10 months at a regional banking client.

Growth in Asia-Pacific accounts for over 10% of Boomi's total revenue, with banking and manufacturing as the primary sectors. Lucas cites the example of a retail chain that automated corporate purchasing with Boomi agents — the agent makes decisions purely based on data, without emotions and without "this is what we've always bought."

What about pricing and availability in the Czech Republic?

Boomi offers a 30-day free trial, followed by paid editions tiered by features. Specific prices vary depending on the scope of deployment and the number of connectors — it is an enterprise solution primarily targeting mid-sized and large companies. For small Czech startups, it may be financially out of reach, but for companies with revenues in the hundreds of millions of crowns dealing with complex integrations between ERP, CRM, and e-commerce platforms, the investment can pay off quickly.

The platform is entirely in English and Czech is not supported. However, this is not a major barrier for technical users (IT administrators, integration specialists) — the interface is designed for a technical audience and uses standard terminology.

In the European context, GDPR compliance is important: Boomi has a dedicated PII agent that monitors data flows and alerts if personal data appears anywhere in violation of the rules. For companies subject to the EU AI Act, this is a practical tool for ensuring compliance.

Agent architecture: it's not just about deployment, but governance

One of the most interesting ideas from Lucas's interview concerns agent architecture — a system that allows companies not just to deploy agents, but above all to understand them. "When an agent decides to hire an employee, you need to be able to explain that decision — for compliance, audits, regulation," says Lucas.

This is a fundamental shift from the current state, where most companies are testing AI agents but few have any idea what exactly the agents are doing, how they make decisions, and whether they violate internal rules. Boomi promises an architecture that will be "agent-aware" — it will know which agents are running in the company, what they are doing, and provide tools for managing them.

For Czech CIOs, this is a topic that will become increasingly relevant in the coming months: as the number of AI tools in companies grows (Copilot, ChatGPT Enterprise, specialized agents), so does the chaos. A platform that ties it all together may make the difference between a successful AI transformation and a regulatory disaster.

The competition isn't sleeping

Boomi is certainly not alone in the market. Workato, MuleSoft (Salesforce), SnapLogic and other integration platforms are also investing in AI. The difference, according to Lucas, lies in the data: "Our competitors rely on recipes that someone wrote because they thought they would taste good. We build on what thousands of organizations have actually done."

The question remains whether customers will value the depth of integration data more than the flexibility and affordability of competing solutions. For companies already using Boomi, however, the expansion into AI agents is a logical step — they don't need to switch platforms, they just gain new capabilities.