Our read
Pro can carry a small production automation: the plan page calls it designed for ongoing production usage. Its documented limits are three users per workspace and US-only hosting, and environment management is listed only under Enterprise, with unlimited users and multi-region deployment. Workato publishes custom roles, a 90-day audit log and eight data regions. How failed runs are surfaced and retried is not described, so test it with a recipe built to fail.
Free, Pro and Enterprise: what changes at each step
Workato describes itself as a platform that combines integration and workflow automation for IT and business teams. Free costs $0 and comes with a one-time grant of 50,000 credits. Pro costs $100 a month. A trial of Enterprise is arranged through sales.
The plan page states that Free and Pro share the same core platform; the difference is credits. Pro adds recurring monthly credits, custom connectors, built-in analytics and core security and access controls, and the plan FAQ describes it as designed for ongoing production usage. What limits it is the list: up to three users per workspace and US hosting. One detail to confirm: the Pro card shows 3.5k credits a month, while the plan FAQ says recurring credits start from 2,500.
Enterprise is the tier that lists unlimited users and workspaces, platform APIs, advanced lifecycle and environment management, enhanced connectivity and multi-region deployment. My reading: as soon as Ops, RevOps and IT all need to build in the same workspace, the three-user cap makes Enterprise a necessity, not an option.
Who can build, approve and deploy
The security documentation describes Automation HQ federation: one control plane where you decide which capabilities and connectors each workspace may use, set usage limits per team, and enforce review and approval workflows on deployments. Custom roles then define who can create, build and deploy workflows, agents and MCP servers. Agent Studio roles separately control test mode and conversation history.
One restriction matters here. The pricing page lists core security and access controls under Pro and enterprise security and governance under Enterprise, but the security page does not map each feature to a plan. Ask which of these roles and approval steps your plan includes before you design around them.
Two further published controls help keep prototype habits out of production: native integrations with AWS Secrets Manager, Azure Key Vault and HashiCorp Vault to avoid plaintext credentials, and approval requests routed to Microsoft Teams or Slack for high-stakes actions. Editorial suggestion: give builders and deployers different roles, so the person who writes a recipe cannot push it live alone.
Dev, test and production: what is published, what is not
The pricing page places advanced lifecycle and environment management in Enterprise, and the security page mentions environment-level and project-level permissions. Neither describes how many environments exist, how a recipe is promoted from one to the next, or how a release is rolled back. Treat those mechanics as open until you have seen them in a trial.
A practical test, offered as an editorial recommendation and not as a vendor feature: during an Enterprise trial, promote one recipe from test to production using an account that is not allowed to edit it, then try to undo the release.
Watching runs: the audit log, its retention, and the error question
Audit logging captures workflow executions, configuration changes and user actions. Retention is configurable up to 90 days, and longer retention is possible through an external SIEM. That is an accountability trail. If your policy requires a year of history, plan the SIEM connection from day one.
Error handling is a different matter. Neither the security page nor the plan page describes failure alerts, retry rules or who gets notified when a run breaks. The homepage presents an assistant called Airo as an IT Ops role that goes through logs, detects, analyzes and resolves issues, but that is a vendor claim, not a description of how failures are handled. Run a recipe that fails on purpose and see what reaches your on-call channel.
Data location, and what you owe your own customers
Workato lists data centers in the U.S., E.U., Japan, Singapore, China, Israel, Australia and Korea. The plan FAQ adds that Pro data is hosted in the United States, while Enterprise customers may choose other locations, subject to availability. The vendor also lists SOC 2 (type not stated), ISO 27001, ISO 27701, ISO 42001, HIPAA, PCI DSS and IRAP; request the reports before relying on any of them.
Several companies name Workato as a sub-processor on their own legal pages (DECLARED). Mindtickle lists it as an integration platform for third-party apps, with data in the United States and email as the data processed. Envoy Global lists it for HRIS integration. Yotpo lists it for data flow integration and automation, with standard contractual clauses for EU and UK transfers. Precisely lists it under data infrastructure. The lesson is practical: if your automations move your customers' personal data, expect to disclose Workato in your own sub-processor list.
What the Atlassian case illustrates, and what it does not
Atlassian's case study, published by Workato (CONFIRMED), describes Project Poly: replacing NetSuite with Oracle Fusion Cloud in nine months. Atlassian reports more than 800 business process changes and 50-plus integrations, about 40,000 finance hours saved, and a financial close in under three days. It also says auditors were impressed by the documentation, reliability and completeness of controls. A separate quote cites over 100,000 manual hours saved across the business, a wider scope than the finance figure.
Read it as an illustration of central orchestration designed for audit readiness. It does not say which plan, how many environments or what error monitoring Atlassian used, so it cannot answer the open questions above.
Method and limits
2026-11-22
This dated snapshot counts distinct companies in public customer evidence. CONFIRMED means a case study, named testimonial or press source; DECLARED means the company itself told usedby; OBSERVED means a vendor listing; SIGNAL means a job post. All four count, but their strength differs. Removed and unlevelled lines are excluded. Examples are a capped selection verified within 90 days. Vendor facts are sourced records checked within 60 days. Coverage is uneven and missing evidence is unknown. These figures measure neither market share nor product performance; capture dates are not adoption dates.