Upskilling Your Workforce for the Agentic Era

AI agents are changing the nature of work. Tasks that once filled hours of a knowledge worker’s day can now be handled by intelligent automation. This shift creates a pressing question: what skills do your people need in an agentic workplace?

The answer is not to abandon existing expertise, but to build new capabilities on top of it. Here is how to approach upskilling for the agentic era.

The skills shift

When agents handle routine work, humans are freed to focus on what agents cannot do - at least not yet. This means greater emphasis on:

• Judgement and decision-making: Handling exceptions, edge cases and situations that require contextual understanding.
• Supervision and quality assurance: Reviewing agent outputs, identifying errors and providing corrective feedback.
• Design and configuration: Defining how agents should behave, what guardrails to apply and how to improve performance over time.
• Stakeholder communication: Explaining agent-driven outcomes to colleagues, customers and regulators.
• Creative problem-solving: Tackling novel challenges that do not fit existing patterns.

These are not entirely new skills, but they take on new importance when agents are part of the team.

Identifying your skill gaps

Every organisation’s skill gaps will be different, depending on the workflows being automated and the current capabilities of the workforce.

To identify gaps:

1. Map the workflows where agents are being introduced or planned.
2. Define the human roles that will remain - supervision, exception handling, design, escalation.
3. Assess current capabilities against those roles, honestly.
4. Prioritise the gaps that will have the biggest impact on success.

Avoid generic lists of “AI skills.” Focus on what your people will actually need to do.

Designing effective upskilling programmes

Traditional training programmes often fail because they are disconnected from real work. For agentic upskilling, take a different approach.

Embed learning in the workflow

The best way to learn how to work with agents is to work with agents. Design training that happens alongside real tasks, with coaching and support available.

For example:

• Pair new users with experienced colleagues during the first weeks of an agent rollout.
• Use sandbox environments where people can experiment without risk.
• Review agent decisions together, discussing why they were right or wrong.

Focus on principles, not just procedures

Procedures become outdated as agents evolve. Principles last longer. Teach people to understand how agents reason, what their limitations are and how to think critically about their outputs.

This builds adaptability. When the agent changes, people can adjust without waiting for new training materials.

Make it continuous

Upskilling is not a one-time event. Agents improve, use cases expand and roles evolve. Build mechanisms for ongoing learning:

• Regular knowledge-sharing sessions.
• Access to updated documentation and resources.
• Feedback loops where people can flag learning needs.

Addressing anxiety and resistance

Some people will see upskilling as a threat rather than an opportunity. They may fear that learning new skills is an admission that their current role is at risk.

Address this directly:

• Be honest about how roles are changing - and what is staying the same.
• Position upskilling as investment in people, not preparation for redundancy.
• Celebrate progress and recognise those who embrace the change.

Resistance often fades when people see colleagues succeeding in the new model.

The role of leadership

Upskilling will not happen by itself. Leaders need to:

• Allocate time and budget for learning, even when it competes with delivery pressures.
• Model curiosity and openness to change.
• Hold managers accountable for developing their teams.
• Recognise and reward new skills, not just old metrics.

Without leadership commitment, upskilling becomes a box-ticking exercise rather than a genuine capability shift.

New career paths

The agentic era creates new career opportunities. Roles like “agent designer,” “workflow supervisor,” and “AI quality analyst” are emerging in forward-thinking organisations.

Consider how your career frameworks need to evolve:

• What new roles are needed?
• How do people progress from user to supervisor to designer?
• What competencies define excellence in an agentic context?

Answering these questions signals to your workforce that there is a future for them - and a reason to invest in developing new skills.

The upskilling imperative

Organisations that neglect upskilling will find themselves with a workforce unprepared for the agentic future. People will struggle, adoption will falter, and the promised benefits of AI agents will remain out of reach.

Those that invest in their people will unlock the full potential of agentic workflows - and build a workforce that can adapt to whatever comes next.