Thoughtworks has launched Agent/works, an enterprise platform intended to help organizations govern, observe and run AI agents across cloud environments. The company describes the product as a control plane and runtime for agentic systems, including autonomous workflow agents and coding-assistant-style tools.
Thoughtworks has launched Agent/works, a platform the company says is designed to help enterprises govern and run AI agents across cloud environments.
In a PR Newswire announcement, Thoughtworks described Agent/works as a platform that gives organizations a “single control plane” and governed runtime for AI agents. The company said the platform is intended to support autonomous workflow agents as well as coding-agent-style tools, reflecting a broader enterprise push to manage AI systems that can take actions, use tools and participate in business processes.
Thoughtworks’ product page positions Agent/works as an “AI agent governance platform” for governing, observing and optimizing agentic workforces. According to Thoughtworks, the platform includes capabilities for policy enforcement, auditability, an agent registry, runtime support and deployment across multiple cloud environments.
The launch is aimed at a practical problem facing companies adopting agentic AI: once AI agents move beyond isolated experiments, organizations need ways to understand what agents are doing, what tools they can access, which policies apply to them and how their behavior is monitored over time.
The Agent/works announcement fits into a wider discussion about how AI governance is expanding from data and machine learning models to agents and tools. Databricks lists a Data + AI Summit session involving Thoughtworks on “extending governance from data to models, agents, and tools” using Unity Catalog and observability.
That framing matters because agentic systems often connect several components: foundation models, application logic, enterprise data, external tools and workflow permissions. Governance therefore becomes less about reviewing a single model in isolation and more about tracking the full operating environment in which an agent acts.
Thoughtworks says Agent/works is meant to address that environment through features such as policy controls, auditability and observability. Those terms are significant for regulated or complex organizations, where AI systems may need to follow internal rules, produce evidence for review and operate consistently across different teams or cloud providers.
Thoughtworks is also emphasizing that Agent/works can fit across multiple cloud environments. The company’s PR Newswire announcement says the platform is intended to govern and run enterprise AI agents “across any cloud,” while the Thoughtworks product page describes a multi-cloud fit.
For large enterprises, this positioning is likely to be relevant because AI adoption is often distributed across business units and technology stacks. A company may have development teams experimenting with coding agents, operations teams testing workflow automation and business teams evaluating task-specific assistants. Without shared controls, those efforts can be difficult to inventory or monitor.
Agent/works is presented as a way to centralize oversight without requiring every agent to live in a single cloud or tooling environment. Thoughtworks describes the platform as including a registry and runtime, suggesting that enterprises can catalog agentic systems and apply operational controls as they are deployed.
The available source material is primarily from Thoughtworks and related event listings, so the claims currently available are vendor descriptions rather than independent performance evaluations. Thoughtworks says Agent/works provides governance, observability, optimization, policy enforcement, auditability, registry functions, runtime capabilities and multi-cloud support.
The company has not, in the provided sources, published independent benchmarks or customer outcome data for Agent/works. As a result, the most supportable interpretation is that Thoughtworks is entering the market for enterprise agent governance with a platform focused on control, monitoring and operational management.
The launch underscores a shift in enterprise AI adoption: organizations are not only asking whether AI agents can complete tasks, but also how those agents can be registered, governed, audited and operated safely within existing technology environments.
Thoughtworks has launched Agent/works, a platform the company says is designed to help enterprises govern and run AI agents across cloud environments.
A control layer for enterprise agents In a PR Newswire announcement, Thoughtworks described Agent/works as a platform that gives organizations a “single control plane” and governed runtime for AI agents.
Thoughtworks’ product page positions Agent/works as an “AI agent governance platform” for governing, observing and optimizing agentic workforces.
Continue reading
