Choosing an iPaaS that can support AI agents is a serious initiative, and third-party research is a reliable place to start. GigaOm’s iPaaS Functionality Evaluation, published September 27, 2026, tested Workato and Boomi across 29 tasks, including six new tests covering agentic automation and the Model Context Protocol (MCP). Workato’s total weighted score was 1,363 against Boomi’s 1,058, which GigaOm describes as a 29% advantage. The widest gaps showed up in the agentic tests. You can read the GigaOm findings and a summary below.
“Workato establishes strong leadership across the board, earning perfect marks in workflow creation, ease of connectivity, and workflow governance, while dominating the Agentic category with standout scores in prebuilt MCP server availability, and human-in-the-loop support.”
Source: GigaOm, iPaaS: Integration Platform-as-a-Service Functionality Evaluation, September 27, 2026
GigaOm iPaaS Functionality Evaluation Criteria
GigaOm scored each platform on functionality, speed of development, and ease of use. Both tools connected to the same Salesforce organization and NetSuite account, so each moved identical data. Six tests focused on how each platform exposes integrations to AI agents: MCP support and general availability, prebuilt MCP servers, exposing an automation as an agent-callable skill, identity and access propagation, human-in-the-loop support, and direct MCP validation.
Workato vs. Boomi: A Comparative Analysis
GigaOm describes Workato as “dominating the Agentic category.”
MCP That Is Ready for Production
GigaOm found Workato’s MCP support generally available and production-grade, with an MCP gateway, a registry of published servers, multi-token authentication, and tool annotations. Workato scored 10 out of 10. Boomi supports the protocol through Agentstudio and a first-party MCP Server connector, but its documentation labels the connector a technology preview not intended for production, and it scored 5.
Prebuilt MCP Servers
Workato offers a managed catalog of more than 100 MCP servers, and an existing Salesforce connection produced a ready-made server with no configuration. GigaOm’s search of Boomi’s marketplace for MCP returned zero templates and four entries, none of them a managed catalog of exposable servers. Workato scored 10, Boomi 4.
From Automation to Agent Tool, With Identity and Approvals
Workato generates the MCP tool schema from a typed recipe function, and attaching it to a server took roughly 10 to 15 minutes. Each caller can present its own token, so an agent acts as a user rather than as the platform, and approvals can run through Slack or Microsoft Teams. GigaOm credited Boomi’s Agent Control Tower as a genuine answer to agent oversight, but Workato still scored higher on all three tests.
Proven End to End
For direct MCP validation, GigaOm connected a standard MCP client to an endpoint, discovered the tool schemas, and sent test payloads with no AI model involved. Workato passed on the first attempt. Boomi could not be tested live because its MCP Server connector was unavailable in the trial account.
Workato vs. Boomi: Field Test Results
- Overall weighted score: Workato 1,363, Boomi 1,058
- Six agentic and MCP tests combined: Workato 370, Boomi 220
- MCP support and general availability: Workato 10, Boomi 5
- Prebuilt MCP servers: Workato 10, Boomi 4
- Direct MCP validation: Workato 9, Boomi 4
Those agentic scores account for 150 of Workato’s 305-point overall lead.
Workato vs. Boomi Comparison: Test Scores and Weights (GigaOm 2026 research)
| TEST | Weight | Workato | Boomi | |
| 1. Agentic | MCP Security & GA status | 10 | 100 | 50 |
| Pre-built MCP server availability | 5 | 50 | 20 | |
| Time to expose automation as skill | 5 | 45 | 30 | |
| Identity/access propagation | 5 | 40 | 30 | |
| Human-in-the-loop support | 10 | 90 | 70 | |
| Direct MCP validation | 5 | 45 | 20 | |
| 2. Management | Setup & Configure Environment | 1 | 9 | 8 |
| Create a Workflow | 10 | 100 | 70 | |
| Ease of Connectivity | 10 | 100 | 70 | |
| Authenticate New Users | 5 | 50 | 50 | |
| Rerun / Restart a Workflow | 5 | 50 | 50 | |
| 3. Automation | Maintain & Revise Workflows | 5 | 50 | 35 |
| Scale Workflows | 1 | 10 | 8 | |
| Scale New Users | 1 | 10 | 10 | |
| Schedule & Execute | 5 | 50 | 40 | |
| Error Handling | 5 | 40 | 45 | |
| 4. Governance | Governance of Users | 5 | 45 | 50 |
| Governance of Workflows | 5 | 50 | 40 | |
| Third-Party Security | 5 | 50 | 35 | |
| API Management | 10 | 80 | 70 | |
| 5. Approval | Document Approval Workflow | 1 | 10 | 7 |
| Security Approval Workflow | 1 | 10 | 6 | |
| Batch Transaction Loads | 5 | 50 | 40 | |
| Real-Time Event Workflows | 5 | 50 | 40 | |
| Send Meeting Notes Doc | 1 | 10 | 5 | |
| 6. Development | Real-Time Error Logging | 1 | 9 | 9 |
| Backup & Restore | 5 | 50 | 40 | |
| DevOps / CI-CD Migrate | 10 | 100 | 100 | |
| Workflows as Code | 1 | 10 | 10 | |
| Grand Total | 1363 | 1058 | ||
Source: GigaOm 2026
Which Platform Should You Choose for MCP and Agentic AI?
According to GigaOm, Workato is the stronger platform for organizations putting MCP and agents into production. Keep the context in view: this is a sponsored report, a point-in-time check of functionality, and Boomi led on error handling and user governance. If your priority is production-ready MCP and fast paths from existing automations to agent tools, Workato is the platform to evaluate.
Book a Workato Demo →
See for yourself why Workato leads on MCP and agentic AI.
Report FAQ
1. What did GigaOm’s iPaaS Functionality Evaluation find about Workato vs Boomi?
GigaOm’s iPaaS Functionality Evaluation, published September 27, 2026 by William McKnight and Jake Dolezal, tested Workato and Boomi across 29 tasks. Workato’s total weighted score was 1,363 and Boomi’s was 1,058, which GigaOm describes as a 29% advantage. The report was commissioned by Workato.
2. How many tests did GigaOm run, and what did they cover?
GigaOm ran 29 tests. Twenty-three were repeated from a 2023 study, and six were new tests covering agentic automation and the Model Context Protocol (MCP). The new tests covered MCP support and general availability, prebuilt MCP servers, exposing an automation as an agent-callable skill, identity and access propagation, human-in-the-loop support, and direct MCP validation.
3. Which platform scored higher on MCP and agentic AI, Workato or Boomi?
Workato scored higher on all six agentic and MCP tests. Their combined weighted scores were 370 for Workato and 220 for Boomi, and GigaOm describes Workato as “dominating the Agentic category.” That 150-point gap accounts for roughly half of Workato’s 305-point overall lead. (The combined totals are our arithmetic from GigaOm’s Table 1, not figures GigaOm printed.)
4. Is Workato’s MCP support production-ready compared with Boomi’s?
GigaOm found Workato’s MCP support generally available and production-grade, with an MCP gateway, a registry of published servers, multi-token authentication, and tool annotations, and scored it 10 out of 10. Boomi supports MCP through Agentstudio and a first-party MCP Server connector, but its documentation labels that connector a technology preview not intended for production or production data. GigaOm scored Boomi 5 out of 10.
5. How many prebuilt MCP servers does Workato offer vs Boomi?
Workato offers a managed catalog of more than 100 MCP servers, and in GigaOm’s test an existing Salesforce connection produced a ready-made server with no configuration. GigaOm’s search of Boomi’s marketplace for MCP returned zero templates and four entries, none of them a managed catalog of exposable servers. Workato scored 10 and Boomi scored 4 on this test.
6. How long does it take to turn an existing automation into an AI agent tool in Workato?
In GigaOm’s testing, exposing an existing Workato recipe as an agent-callable MCP tool took roughly 10 to 15 minutes end to end. Workato generates the tool schema from a typed recipe function. Each caller can present its own token, so an agent acts as a user rather than as the platform. Workato scored 9 and Boomi scored 6.
