In 1965, Stevie Wonder walked into a studio to record “Uptight (Everything’s Alright).” The job of producing that record was about pretty much the same things every record before it had been: get the right musicians in the room, capture the right takes, mix it down. The producer’s value was inseparable from their ability to coordinate the right people at the right moment.
But by 1972, the realities of production had changed radically. Wonder was recording “Music of My Mind” largely alone, playing almost every instrument himself, including the TONTO synthesizer, a machine that could produce sounds no human being had ever made. He was then layering every recording together to create a whole that was much greater than the sum of its parts.
In those few years, the music producer’s role hadn’t disappeared, but the reality of what being a producer meant had completely shifted to adapt to the times. Thanks to the evolution of technology, you could now turn one person’s expertise into a whole hybrid human/machine symphony.
Some of you may be wondering why I’m opening this article in a Resource Management Institute publication with a music history lesson. And some of you may have already inferred that this is actually a pretty close analog for the journey the resource management profession is currently on, and how fast what’s possible and what’s expected can change.
The question of “who’s available?” has been the logistical center of gravity for RM for decades. It’s not going away overnight! But it is becoming the wrong thing to ask. With AI and the pressures, expectations, and opportunities it creates, the field is being reconstituted around harder, more consequential questions:
- Who delivers outcomes?
- What team combinations actually work?
- What does “capacity” mean when part of your workforce isn’t human?
At Kantata, we’ve been doing the work to uncover just what the future of resource management looks like, drawing on the insights of those in the field and from RMI research on AI’s impact on the field to answer one primary question: if we fast forward to June 2028, what will you say your reality is as a resource management practitioner?
Five realities emerged.
Reality #1: I’m orchestrating a hybrid workforce of humans and AI agents
This one is happening whether we’re ready or not. According to Kantata State of the Professional Services Industry research, eighty-seven percent of professional services leaders say they’re preparing to manage AI agents as part of their delivery workforce. And 90% say their systems will need to attribute work and costs to both humans and agents. The organizational expectation is set.
The RM community is not yet where it needs to be. In the RMI’s Resource Management in the Age of AI report, 52.7% of resource managers say they’re not equipped to orchestrate hybrid teams. And the most cited barrier, named by nearly half of respondents in the RMI survey, isn’t technology access or budget. It’s limited understanding of where AI should even apply.
That’s a capability gap, not a tools gap. The work ahead will move past adding agents to the roster towards developing the ability to know when and where human-agent collaboration generates better outcomes than either working alone.
Reality #2: We staff based on proof of outcome
Do you actually know which team combinations, in which types of engagements, consistently deliver strong results? For most organizations, the honest answer is no.
When surveyed, 73.4% percent of resource managers say it would be very or extremely valuable to know which team combinations consistently deliver strong outcomes. Only 7.3% say that data is easily accessible and routinely used in staffing decisions. And for nearly one in four resource managers (24.6%), outcome data simply isn’t visible to them at all.
This isn’t just dependent on whether your organization is shifting to outcome-based pricing (for many organizations, true outcome-based pricing isn;t in the cards, in spite of what industry hype says.). As the perceived value of the billable hour goes down, proof of outcome will become the currency of the future, even if you’re not holding revenue in contingency until an outcome is provable. More than half of resource managers already say that demonstrated outcome experience influences staffing recommendations for prospective deals. The instinct is there. The infrastructure to act on it systematically is not.
The 2028 reality isn’t that we’ll suddenly know everything. It’s that we’ll have stopped accepting “we don’t have that data” as an acceptable answer.
Reality #3: We optimize for revenue per head as our primary KPI
Utilization made sense as a primary metric when scaling revenue meant scaling headcount. But that equation doesn’t work as well for where RM is at now… and where it’s headed. According to 89% of PS leaders surveyed by Kantata, future revenue growth will depend more on scaling AI than scaling human resources.
This doesn’t mean utilization will disappear from the dashboard. It just won’t be the top north star number leadership cares about anymore. The KPI is shifting from “are my people busy?” to “are my people generating value?”
Revenue per head forces a different set of questions. It asks whether the deployment of a given person, in this role, on this project, with this team structure, is the highest-value use of that capability.
It’s a harder question, but it’s the right one.
Reality #4: Our skills are a living system, not a static list
Sixty-eight percent of resource managers want AI to infer skills from work actually being performed, not from manually maintained profiles that went stale the moment they were filled out.
This is the difference between skills as HR record-keeping and skills as operational infrastructure. A living skills system surfaces adjacent capabilities. It tells you that the person who spent six months leading a data migration probably knows something useful about change management, even if neither word appears in their profile. It identifies emerging capabilities before anyone thinks to go update a form or spreadsheet.
The organizations that get to 2028 with a functional skills infrastructure will have a staffing intelligence advantage that compounds. The ones still running on self-reported profiles and annual review cycles risk being left behind.
Reality #5: AI does the scheduling. I analyze, validate, and advise
This is the reality that provokes the most conversation and, I think, the most anxiety.
It’s not that resource managers are against AI streamlining the scheduling process – 66% of resource managers want AI to handle automated skills matching and scheduling, while 54% want it to recommend optimal human-agent team combinations. That’s a clear signal from the field that the appetite is there.
But the biggest question isn’t really about whether anyone wants AI to take over scheduling. It’s about whether CEOs, boards, and senior leaders will expect AI to take over scheduling. And, by 2028, they will (with human-in-the-loop caveats), which means they’ll ask what the RM function is contributing if an agent is making the actual assignments.
Automated resource scheduling has been technically feasible for years. GenAI didn’t invent the possibility, though most organizations have been slow to hand over staffing workflows to advanced algorithms because of other gaps like lags in manual skills maintenance. What’s different now is the executive expectation layer. The 2028 RM who hasn’t answered the question “what do I do that AI can’t?” is in a precarious position.
The answer, when articulated well, is actually a powerful one: analysis, judgment, organizational context, stakeholder management, and the ability to translate workforce strategy into executive-level guidance. That’s not a narrow slice by any means! It’s the core of a genuinely elevated function. But it requires actively building toward that role, not waiting to see what’s left.
Bridging the Gap Between Now and Then
What I find most striking across these five realities is that none of them require imagining some distant, implausible future. Most resource managers I talk to believe they’re coming. And many believe they’re already partially here.
The vision is there. What’s hard is the distance between the operating model most organizations are running today and the infrastructure required to make these realities work.
Because the ultimate reality is this: You cannot run a 2028 playbook on 2018 infrastructure. And too many organizations are about to try to do just that if they can’t strategically redirect towards what resource management will quickly need to become. When I see 72% of resource managers report that the resource management function in their business is primarily (or solely) operational, with limited strategic influence, I know we have a long way to go in a short amount of time.
The five realities laid out here are achievable. The distance from here to there is a function of decisions being made right now: about systems, about data, and about what we’re actually trying to measure – and why.
The producers who thrived in the early seventies weren’t the ones who resisted synthesizers. They were the ones who understood what synthesizers made possible and built new creative frameworks around them before everyone else did.
Seven years is all it took to completely transform a producer’s role. The clock on resource management is already running.