Workforce planning used to mean one thing: how many people to hire, where, and when. In 2026, a growing number of HR functions are being asked to plan capacity across two categories of workers at once, human employees and AI agents, and Atlassian has become the clearest evidence yet that this split is turning into a formal HR discipline rather than a side project.
A new title enters the org chart
In a blog post published on Atlassian’s “Inside Atlassian” site, Avani Prabhakar, the company’s chief people and AI enablement officer, laid out five beliefs reshaping how the company runs HR. Ranked first: HR must evolve from a function that manages headcount into what Prabhakar calls an organizational “capacity architect,” balancing human talent against AI agents rather than treating agents as a separate IT concern.
That belief is not just a stated principle. Atlassian has created a role, Director of Capacity Planning, Human and AI, Strategic Modeling, whose job is to build the frameworks that decide which work goes to people and which goes to software. Alicia Lenart, Atlassian’s vice president of HR business partners, described the logic behind the title directly: “Capacity is two things, right? It’s the human folks that you have, but it’s also the agents that you have.”
Why one company’s org chart is a signal, not a curiosity
A single title at a single software company would ordinarily be a footnote. What makes Atlassian’s move notable is how closely it tracks a labor-market shift already visible in harder numbers. PwC’s 2026 AI Jobs Barometer found that entry-level roles in occupations with high AI exposure are now seven times more likely to require senior-level skills, such as leadership, than comparable roles were before. Roles that AI enhances rather than replaces have grown 39% since 2018, more than double the 17% growth rate for roles where AI simply automates tasks.
Put together, the data and the Atlassian appointment describe the same underlying problem: the line between “a person’s job” and “a task an agent can do” is moving fast enough that organizations can no longer treat it as settled once a year during headcount planning. Atlassian’s answer is to give someone a standing mandate to keep redrawing that line, continuously, as part of HR rather than IT or finance.
The other four beliefs matter almost as much
Prabhakar’s post frames the capacity-architect role as one piece of a broader redesign. HR, she argues, should organize around end-to-end employee-journey outcomes instead of functional silos, such as recruiting, compensation, and learning, operating in isolation. Employee data should shift from static administrative records toward “connected context” that lets AI systems reason about a person’s situation rather than just store it. Talent programs need to track fluid skills rather than fixed roles, since annual performance cycles were built for a world where jobs changed slowly and now do not reflect how people actually work alongside agents week to week.
The fifth belief is the one meant to reassure HR practitioners rather than unsettle them: relationship-based HR work does not disappear. “An AI can brief a manager on their team’s data,” Prabhakar writes, “but it can’t carry forward the history of a coaching relationship.” The capacity-planning function she describes is explicitly designed to protect that human work by taking mechanical allocation decisions off managers’ plates, not by eliminating the managers.
What the capacity-planning shift means for the HR leader
For HR leaders outside the small group of companies already experimenting with this, three implications follow.
First, workforce planning needs an owner who sits inside HR and understands both headcount economics and what AI agents can reliably do, rather than splitting that judgment between an HRBP and an IT or automation team that rarely talks to HR about org design. Without that owner, capacity decisions default to whichever team moves fastest, which tends to produce inconsistent, ad hoc automation rather than a coherent plan.
Second, the skills data from PwC suggests the redesign work cannot wait for a formal reorg. If entry-level roles in AI-exposed functions already demand senior-level judgment, career ladders and job architecture built around gradual skill accumulation are already out of date in those functions. HR teams that have not audited which roles fall into that category are working from stale assumptions about what junior hires need to know on day one.
Third, the shift changes what HR technology needs to track. Systems built to manage headcount and requisitions were not designed to model a mixed workforce of people and agents with different cost structures, capacity limits, and failure modes. Vendors serving this market are likely to face pressure to add agent-capacity modeling alongside traditional workforce planning tools, echoing the reskilling infrastructure investment already forming elsewhere in the market, including a $1 billion coalition of AI developers building retraining programs for workers displaced by the same automation shift.
What to do next
HR leaders do not need to copy Atlassian’s exact title to act on the belief behind it. The practical starting point is smaller: identify which roles in the organization already blend human and AI-agent work in an ad hoc way, and assign explicit ownership, inside HR, for deciding how that split should evolve over the next planning cycle. Waiting for the reorg to arrive on its own risks leaving that decision to whichever team accumulates the most agent licenses first, which is rarely the outcome a deliberate workforce strategy would choose.
Source: Atlassian
