Atlassian (Nasdaq: TEAM) is expanding a Forward Deployed Engineering program that places applied AI engineers alongside enterprise customer teams, a delivery model designed to move selected AI projects from experimentation into production. The company said it has 46 Forward Deployed Engineers today and plans to grow the group toward 100 by its third fiscal quarter, with staff working in the United States, EMEA and India.
The program targets a persistent enterprise problem: AI initiatives often produce demonstrations before they produce durable workflow changes. Atlassian said its engineers work inside a customer environment, begin with a business outcome and then build production code, agents, system connections and redesigned workflows. For IT leaders, the immediate benefit is a more direct route from a defined operational problem to a deployable implementation, rather than a broad AI assessment that leaves integration and operating details unresolved.
Embedded Engineers Turn AI Ambitions Into Specific Workflows
Atlassian said the team has worked with more than 100 enterprise customers and helped put more than 80 AI agents into production. The company positions the role as distinct from advisory consulting: engineers identify one or two high-value use cases with the customer, then build against the systems, knowledge, permissions and handoffs involved in the work. The program is available to a select group of enterprise Cloud customers using Rovo, Atlassian said.
That distinction matters because an enterprise agent is not only a model choice. It also needs access to the right information, controls over what it can do and a place in an existing business process. Atlassian says its Forward Deployed Engineers evaluate who participates in the work, where the needed knowledge resides, where handoffs fail and what permissions an agent requires before building the solution. The company says teams typically become self-sufficient in about 12 weeks, although that is a company-reported measure rather than a guarantee for every deployment.
The Model Makes Governance and Reuse Delivery Priorities
Atlassian’s published playbook emphasizes that permissions, security and auditability should be part of the architecture from the outset, with autonomy increasing as trust grows. It also calls for measuring AI against the work it replaces and reusing context where possible. Those are practical requirements for CIOs assessing whether an agent can move beyond a pilot without creating a new governance or cost problem.
The company also says its field engineers feed lessons from customer deployments back to product teams. That creates a potential multiplier: a difficult integration or workflow pattern found at one customer may inform improvements that other users can apply on the Atlassian platform. The claim is strategically important, but its value will depend on how consistently Atlassian can translate bespoke work into repeatable product capabilities.
Why This Shifts the Enterprise AI Delivery Equation
Legacy software rollouts often separate implementation services from product development, while many AI pilots separate model experimentation from the business process that must ultimately change. Atlassian’s approach combines those activities around a limited set of measurable use cases. It does not remove the hard work of data access, integration or change management, but it puts engineers accountable for those constraints closer to the customer team. That makes the program a competitive bet on execution: enterprises that can turn validated use cases into governed, reusable production workflows will gain more from AI than those that remain at the pilot stage.