Building the Future of Organisational Learning One Agent at a Time

How can L&D teams explore agentic tools while staying grounded in learner needs?

Agentic AI is not only about automating basic process work. It can also be used to augment, support and enhance the work that specialist teams and functions perform within organisations in a way that can broaden their scope and increase agility.

Learning is such an important enabler for enterprise AI adoption, and the change management that it requires. One thing AI has taught us is how learning is now an attribute of intelligent systems, and not just an occasional activity that happens away from the context of our daily work. We finally have the tools to create the learning organisation that Peter Senge and others have written about for so long.

For years, Learning & Development (L&D) teams have juggled multiple mandates: aligning to business priorities, building talent pipelines, ensuring regulatory compliance, enabling transformation, and responding to shifting employee expectations around learning and growth. But now, with the rise of enterprise AI and its adoption and adaptation challenges, learning is front and centre of organisational transformation. L&D functions need to adapt to the changing scope, fluidity and speed of learning that is needed today.

There will of course be those who try to take the basic processes of corporate learning and automate it, with AI-generated courses, AI coaches and so on. But rather than using agentic AI to automate or replace learning roles, could we instead deploy AI to augment and enhance the work of L&D teams to help meet the challenge of upskilling the organisation in a time of rapid change?

Some of the challenges L&D teams face today relate to how slow their processes are to complete a cycle of analysis → design → delivery compared to the speed at which new technology is changing the way we work. Previously, learning needs analysis might be done periodically, with a gap of a year or more between identifying a need, planning a programme and delivering it.

But if practitioners are supported by agents working continuously in the background to scan signals, spot patterns, assemble options, and prompt action, agentic AI could potentially help L&D teams to scale up and accelerate their most important activities:

• Learning Needs Analysis becomes a live, responsive process, with agents continuously mapping capability demand, surfacing skill gaps, and feeding into workforce planning.

• Programme Design becomes more modular and tailored, as agents help remix and personalise learning journeys to support leadership, technical, or soft skills development. New content can be added into the mix more quickly, building on existing foundations.

• Digital Learning Services become smarter, as agents help maintain LMS/LXP ecosystems, ingest and curate external content, and personalise delivery across formats.

• Performance Support tools become more contextual, with agents offering job aids, playbooks, or community resources at the moment of need, rather than assuming people will search for and find their content.

• Coaching, Mentoring & Career Development become more personalised and scalable, with agents matching participants to mentors, suggesting reflection prompts, and monitoring engagement, while helping employees explore internal pathways, identify skill gaps, and assemble personal learning plans.

These are not distant possibilities. The pieces already exist but are fragmented across platforms, repositories, frameworks, and human expertise. The agent layer could be a missing piece of the puzzle that helps tie it all together. The goal here is to help L&D run a system that works and adapts continuously, rather than following the old rhythm.

In today’s edition, I will dig into one agent type that we think will be increasingly important as part of the L&D technical stack: a personal learning companion that sits beside the learner and helps them find their next best step. We believe this will play a key role in learning needs analysis, learning delivery, and also the personalisation of learning experience. And by starting at the learning experience interface with the customer, it could also help inform and accelerate back-end learning system improvement

Building the Future of Organisational Learning One Agent at a Time

If the value of L&D lies in connecting business needs, content, and learner experience, then agentic AI could help extend its reach and reduce the lag between learning needs analysis and delivery. Agents can scan signals, assemble journeys, and surface insights, giving professionals more time to focus on strategy, quality, and learner support.

We’re already seeing signs of this shift in three promising areas:

1. Continuous Capability & Skills Sensing

   Traditional skills surveys and competency frameworks go out of date almost as soon as they’re published. Agentic systems offer a live, responsive alternative, drawing from work outputs, project data, and learning activity to help L&D and business partners spot emerging gaps and shifting priorities to maintain real-time capability and skills maps.

2. Curation and Assembly of Learning Experiences

   The problem isn’t content volume, it’s accessibility. Agents can trawl repositories, filter for relevance, and recommend the right mix of articles, videos, simulations, or exercises, all matched to a learner’s goal, context, or role.

3. Business Partner Support and Advice

   L&D business partners often spend hours preparing insights to guide line leaders. Agents can automate much of that groundwork: identifying trends, highlighting capability risks, and preparing draft recommendations, freeing human partners to focus on higher-value conversations and change enablement.

In each case, the human role doesn’t disappear, but it evolves. L&D professionals move from managing delivery to orchestrating systems - from pushing content to enabling personalised discovery, and from static planning to real-time responsiveness.

And the most immediate, visible, and learner-centred place to start is with a personal learning companion.

Opening a New Space for Personalised Learning

Among the many ways agents could support L&D, the Personal Learning Agent feels like one of the most promising starting points. It offers a glimpse of a different way for people to engage with learning, more immediate, more tailored, less dependent on formal courses.

You can imagine it as a kind of learning companion or concierge.

A learner might begin with a simple conversation: “I want to strengthen my stakeholder-management skills”, and the agent could propose a mix of existing resources: a few articles, a podcast episode, perhaps a short simulation exercise.

Over time it could adapt suggestions to the learner’s context and preferred formats, and offer small nudges or reflections in the flow of work, as well as acting as a search agent that supplies a user with articles, links and content related to their learning goals.

What makes this interesting is not that it replaces L&D’s work, but that it shifts the focus towards curation and orchestration.

The agent can do the first pass, finding, assembling, and presenting options, while L&D teams set the standards, decide which sources to trust, and guide how these experiences fit into a broader capability agenda.

Used in this way, a personal learning agent might also surface useful signals:

• what kinds of goals employees are setting,

• which assets they return to,

• where they struggle or disengage.

Those signals could help L&D refine both its content repositories and its wider development priorities.

We do not yet know exactly what this capability will look like in every organisation, and that uncertainty is part of its value. Exploring personal learning agents offers a chance to open up a new space for learning: one that is more conversational, data-informed, and learner-driven, yet still guided by human judgement and oversight.

Mapping the Building Blocks

If we think of the Personal Learning Agent as an emerging capability for L&D to build and steward, we can begin to sketch out the building blocks that make it possible.

This is not a final blueprint — more a first map of the territory that L&D leaders might explore.

At a high level, five areas seem to matter most:

1. Core Systems: the platforms and repositories the agent draws from: the LMS/LXP, curated internal content, external learning libraries, and secure access to knowledge assets.

2. Data Foundations: signals about learners and work: profiles, skills frameworks, usage data, feedback, sometimes even performance data (handled carefully and ethically).

3. Agentic Software: the orchestration layer that brings it to life, search and retrieval, summarisation, recommendation, scheduling and nudging. In most cases this will be a mix of off-the-shelf co-pilots and L&D-specific agentic tools.

4. Services & Processes: the human scaffolding: oversight of content quality, standards for sources, guidance on ethical use of learner data, support for continuous improvement.

5. People & Skills: new competencies for the L&D team itself. From prompt-as-requirements design to interpreting agent analytics, and curating experiences rather than producing all content in-house.

These building blocks can be assembled in many ways.

Some organisations may already have strong data foundations but limited curated assets; others may need to focus first on tagging and cleaning up their existing content.

The map is therefore an invitation to explore what is already in place and what needs attention next.

Key Applications of the Personal Learning Agent

A personal learning agent is not a single tool but a way of weaving AI into the fabric of learning.

Looking across typical L&D services, a few early applications stand out.

Each builds on the same basic idea: a conversational companion that draws from curated assets, but emphasises a different aspect of the learner experience.

1. Goal-Driven Journey Assembly

   Instead of searching catalogues or waiting for a course to be scheduled, a learner can set a goal in plain language: “I need to lead my first cross-functional project”, and the agent proposes a draft journey using approved assets and activities.

2. Adaptive, Multi-Format Delivery

   The agent can present material in the format that best suits the learner’s situation: a quick video or job aid at the moment of need, a podcast summary for the commute, a reflective prompt after trying a new skill.

3. In-Flow Guidance and Nudges

   Because it can connect to work platforms, the agent can surface a tip or micro-exercise just when it is most relevant, bridging the gap between formal learning and daily practice.

4. Feedback-Rich Learning Loops

   As learners interact, the agent gathers lightweight data on what they seek, what they skip, and where they stall. This creates a feedback loop that helps the agent improve recommendations and gives L&D valuable insight into emerging capability needs.

5. Team-Level Insight and Support

   Aggregated (and privacy-respecting) signals from personal agents can reveal patterns, such as a spike in requests around a new regulation, allowing L&D to respond faster at the organisational level.

These applications hint at how the capability can connect the learner’s personal development journey to the organisation’s wider capability agenda, without imposing more formal programmes.

Experiencing the Agent in the Flow of Work

Imagine being a product manager preparing for a new market launch.

You’ve heard about the company’s new learning companion built into your collaboration platform, and you open it between meetings.

You: “I need to get better at framing customer problems for our next discovery sprint.”

Agent: “Got it. Would you like a quick set of resources to start today, or a more structured 3-week journey?”

You pick the quick-start option.

The agent proposes a few items already approved by L&D:

• a short article on problem-framing,

• a 10-minute podcast from a respected product leader, and

• a practical exercise you can run in tomorrow’s team session.

It offers to send a reminder and reflection prompt after you’ve tried the exercise.

Later that week, as you upload notes from a stakeholder interview, the agent pops up gently:

“Would you like a quick debrief template to structure your insights?”

It feels less like a course and more like a helpful guide sitting alongside your work, surfacing relevant assets, nudging reflection, and quietly logging progress in your learning record.

Behind the scenes, L&D reviews anonymised patterns: they see that many people preparing for launches are requesting help on customer discovery, prompting them to tag more assets and perhaps create a focused learning sprint.

Building the Capability

For most organisations, the personal learning agent won’t arrive fully formed. It will emerge, incrementally, from a set of foundations that L&D teams already have, and from a willingness to explore new roles, workflows, and partnerships.

To treat this as a capability is to see it not as a one-off tool, but as an evolving combination of systems, standards, practices, and people that together enable a new kind of learning experience.

Several practical steps can help L&D teams begin. Read on for ideas on how to get started.