Atomicwork, an IT service management startup founded in 2022, has opened an AI Coworker Marketplace of prebuilt AI workers for enterprise IT and security teams. The launch, announced on 6 November, offers roles that are certified before deployment, meaning each has been checked for the integrations and skill coverage its job requires.
The first lineup has an enterprise operations coworker alongside CloudOps engineer, AccessOps manager, SecOps manager and DeviceOps manager roles. They work across AWS, Microsoft Azure, Google Cloud Platform, Okta, Active Directory, Splunk and Google SecOps. Companies can tailor each one to their own infrastructure and business logic, while administrators decide which systems an agent may reach and which actions it may take.
Atomicwork says access automation is where the gain shows: requests for apps needing no approval resolve in a median of about 18 seconds, and about 2.5 hours where approval is required. The industry norm it cites is a day to three business days. Its earlier AI Workforce launch framed coworkers as having an identity, a service account, goals and token budgets, each answering to a human manager.
About Atomicwork
Atomicwork is an enterprise AI company, founded in 2022, that builds an agentic IT service management platform in which AI agents resolve employee tech issues and automate routine support. It is led by CEO Vijay Rayapati, who earlier founded the cloud-management firm Minjar, later bought by Nutanix; its other co-founders, Kiran Darisi and Parsuram Vijayasankar, come from the founding team of Freshworks.
The company announced a Series A of over $25 million on 28 January 2025, co-led by Khosla Ventures and Z47, with Battery Ventures and Peak XV Partners also taking part. In September 2025 it took a strategic investment from Okta Ventures, saying it had raised over $40 million across its seed and Series A rounds.
CEO Vijay Rayapati argues that most AI-native ITSM vendors offer little evidence their agents work, and that a marketplace listing only claims it, which has left enterprises wary. Atomicwork wants verified coverage of the systems an agent touches, tested on real enterprise use cases. It calls the launch a first step toward a third-party marketplace where partners publish coworkers and certify them against its enterprise bench, adding skills through its Model Context Protocol store and skill registry.
Sources · 2