Make learning easier to approach
Important policies can feel heavy. Short explanations, visuals and manageable steps offer a more approachable experience.
Working Product POC V1
AI Employee Companion
Helping employees find their feet,
without putting work on hold.
A working product experiment exploring how contextual AI could bring workplace support, learning and practice into one continuous employee experience.
Demo presented at HRFest2026 and SWDA Employees Connect+
THE COMPANION IN ACTION00:29
Read the video transcript
An employee starts complaint-handling learning and asks for an example of empathy. The Companion explains how to acknowledge a customer’s frustration. The employee switches to an annual-leave question. The Companion explains that full-time employees receive 14 days each year and requests go through the HR portal with supervisor approval. When asked “Can we continue?”, it recalls the Empathise step and introduces Apologise. This demonstrates continuity within the same conversation.
Different organisations.
Similar human problems.
Across roles and industries, I kept encountering a similar challenge: how do we help people learn, change and perform without making the experience another burden?
Bring support closer to work
For frontline teams, learning competes with customer and operational demands. Guidance needs to fit the working day.
Give people room to practise
People may hesitate to admit uncertainty. A lower-risk space to ask and rehearse could help them prepare for workplace situations.
Experience behind these priorities
Enterprise compliance learning, regional frontline digital learning and work supporting cultural and behavioural change informed these priorities. The recurring human problems shaped the hypothesis across different organisational contexts.
THE HYPOTHESISWhat if support, learning and practice could happen in one continuous experience, without losing the thread when an employee’s needs change?
Different needs. One companion.
Employees do not always need a course. Sometimes they need an answer, a refresher or a chance to practise.
New to the role
“Help me get up to speed.”
A question during work
“I need an answer now.”
A quick refresher
“Remind me how this works.”
Preparing to act
“Let me try this before I do it for real.”
An interruption
“Let me return to where I stopped.”
Choose the format that fits the moment
Ask
Explain it to me.
An answer with room for follow-up questions.Watch
Show me quickly.
View the learning video ↗Read
Give me the essentials.
View the LEARN visual guide ↗Listen
Help me revisit the idea.
Audio reinforcement: a format to explore.Practise
Let me try it safely.
A workplace scenario with guidance and feedback.The design starts with the employee’s moment of need. Audio is an exploration, not a demonstrated feature of this version.
Switch tasks without losing your place.
Work rarely happens in one uninterrupted sequence. The tested conversation moved from learning to an HR question, then back to the previous learning topic.
- 01 / LEARN
“Explain empathy with an example.”
- 02 / ASK
“How do I apply for annual leave?”
- 03 / GET SUPPORT
Guidance on the HR portal and supervisor approval.
- 04 / RETURN
“Can we continue?”
Back to the empathy topic.
Shortened excerpts from tested exchanges. See the working demo above.
What this demonstrates—and what remains open
The scenario demonstrates continuity within one open conversation. Reliability across different questions and learning journeys still needs testing. Progress currently resets on refresh.
A workplace pilot would need to establish whether this improves understanding, reduces repeated questions or helps employees apply what they learn.
The learning question: completion versus understanding
Compliance learning helped shape this experiment. Finishing a course and knowing what to do in a workplace situation are different kinds of evidence.
A familiar endpoint
“I completed the course.”
Completion shows that someone finished the activity.The question I want to explore
“Can I apply this at work?”
Questions and practice could reveal where further support is useful.More responsive learning is an experimental direction. This POC does not establish improved learning outcomes or demonstrate a complete adaptive-learning engine.
Meet Sarah
Follow a fictional new employee as she gets started, learns and practises. Her journey illustrates the experience behind the Companion.
ILLUSTRATIVE EMPLOYEE JOURNEY01:21
Read the journey summary
Sarah’s journey moves from a welcome and getting started to learning with an AI companion. She can ask questions and practise as she prepares for work.
From workplace questions to a working experiment.
I designed and built the Companion to make these ideas tangible enough to test.
Discover
Identify recurring workplace problems and form a hypothesis.
Design
Shape the experience across support, learning and interruption.
Build
Create a responsive prototype that people can try.
Test
Review scenario responses and identify what needs further work.
WHAT I’VE LEARNED SO FAR
A useful answer is only part of the experience. Helping someone continue after a change of topic matters too.
How do we know what support is actually needed?
My second POC explores diagnosing the workplace need before choosing an intervention. The appropriate response might be learning, practice, performance support—or a change that does not involve training.
A separate exploration. Not an additional feature of the current Companion.Let’s connect
This project brings together my workplace experience, learning design and AI prototyping. I welcome conversations about roles, pilots and project partnerships.
Explore the Prototype