Service Innovation in the Age of Continuous Upskilling

By Andrea Biancini

There is a quiet but consequential shift happening inside organizations that deliver services at scale. Learning is no longer something that happens before service delivery. It is something that happens through it.

The half-life of professional skills has contracted so dramatically that the episodic upskilling model, with periodic retraining events followed by long stable plateaus, has simply broken down. Organizations still operating under that logic are accumulating invisible debt: a growing gap between what their people know and what the work actually requires. The question for service science is not whether this is happening. It is: what does it mean for how we design, deliver, and govern services?

From Event to Fabric

The organizations responding most effectively are not simply accelerating their training calendars. They are redesigning service delivery itself so that learning is embedded in the act of service. The distinction matters. Accelerated episodic training produces better-prepared workers who then face a deteriorating environment. Embedded learning produces organizations that improve through operation, systems that do not just deliver services but generate, absorb, and distribute knowledge as a byproduct of doing so.
In service science terms, this is the difference between a service system that consumes human capital and one that continuously replenishes it.

Organizations as Learning Systems

The conceptual move required here is significant. We are accustomed to thinking of organizations as service delivery mechanisms that also happen to train their people. The emerging model inverts this. The organization is primarily a learning system, and service delivery is the primary mechanism through which learning occurs.

If the service interaction is also a learning event, then service architecture must account for both dimensions simultaneously. We are no longer only asking: how do we ensure consistent quality of output? We are also asking: how do we design service processes that make the people delivering them more capable by virtue of having delivered them? Conventional KPIs (resolution rates, satisfaction scores, efficiency ratios) do not capture competence growth. The organizations that solve this measurement problem will have a structural advantage that compounds over time.

AI: Augmentation with Accountability

AI is simultaneously the primary driver of skill obsolescence and the most powerful tool available for embedding learning into service delivery. A well-designed AI-augmented service environment does not simply replace human judgment. It surfaces the reasoning behind decisions, makes tacit knowledge explicit, and creates structured opportunities for practitioners to interrogate and internalize what the system is doing. The training is, in effect, the process.

But this works only if human practitioners retain genuine agency, if they are actively engaging with AI outputs rather than simply ratifying them. Organizations that design AI-augmented services to bypass rather than develop human judgment are not building learning systems. They are building fragile systems that will fail precisely when novel situations demand the kind of judgment that only experience and continuous learning can produce.

Upskilling as a Service Science Problem

Approaching continuous upskilling through the lens of service science means, first, recognizing that upskilling itself is a service — one with value co-creation dynamics, quality dimensions, and systemic interdependencies. The person being reskilled is an active co-producer of their own competence development, not a passive recipient.

It means, second, taking lifelong learning seriously as a service design problem. Most learning architectures follow an educational logic: structured curricula, defined outcomes, fixed assessment points. This is mismatched to the continuous upskilling condition, which is non-linear, context-dependent, and driven by emergent needs. Service science, with its concepts of modular architectures, dynamic value constellations, and platform-mediated co-production, has the toolkit to address this.

This approach aligns with emerging evidence that Project-Based Learning (PBL) and Learning-by-Doing (LBD) are the most effective mechanisms for competence development in dynamic environments. As routine work becomes automated, all remaining work becomes inherently project-based and unpredictable. Service systems that embed learning through real-world problem-solving position practitioners to thrive in this reality. The design of service delivery processes themselves becomes a learning design problem.
It means, third, connecting individual learning to organizational capability. Continuous upskilling at the practitioner level is necessary but insufficient. The organizations that sustain competitive advantage are those that convert what any individual learns into something rapidly available to the collective.

A Moment for the Field

ISSIP’s upcoming IVI Summit arrives at the right moment for this conversation. The questions worth bringing are not primarily technical: they are design questions. How do we build service systems that learn? How do we measure competence growth alongside output quality? How do we deploy AI in ways that augment rather than bypass human judgment?

The age of continuous upskilling is not coming. For many organizations, it is already here. The question is whether the service systems we design will be adequate to it.