Arrived from Wordology? You're in the right place. Antelope Camel is our new name: same company, same standards, same person.
Antelope Camel

// the human your AI needs

Where you need a human in your AI mix

AI makes learning fast. On the topics you can't afford to get wrong, fast can do harm.

// any AI-made learning

Where AI needs a human

AI is great at getting learning projects up and running. But learning designers aren't extinct (yet!).

[01]

Whether a scenario will retraumatise someone

People bring their whole lives with them when they step into a training experience. On sensitive topics, it’s almost certain someone taking part has Lived Experience. AI can’t tell when a scenario lands too close to home. We build for that person from the outset.

[02]

Whether a character is culturally real and safe

AI carries the racial and cultural bias of what it learned from. Our characters come from consultation and Lived Experience, not unchecked stereotypes.

[03]

Whether it's accurate for your context

AI fills gaps with confident guesses. Your policies, processes and people aren't always in what it learned from. We check it against how your organisation actually works, so it’s right for your people, not just plausible.

[04]

Whether it will change what people do

AI can produce impressive content fast. Content is one piece of the learning design puzzle. Changing behaviour takes knowing your learners, your context and then your content.

// AI role-play

Using AI role-play?

You bring the tool. We write what happens inside it, the same way we write any scenario: starting with the person.

simulation.designhuman-written
character: a real person, with a real need
disclosure: what they open up, and when
rubric: fair and achievable
guardrails: safe for sensitive conversations
status: tested until it behaves ✓

Whether the scoring is fair

AI tends to reward the right words, not the right judgement. Our rubrics score what learners do, so different approaches can succeed.

When a conversation should stop

AI will follow a conversation down almost any path, and not every path leads where your learners need to go. We write guardrails that keep the character on track, and end the conversation safely when it needs to stop.

Here's one.

AI role-play / demo

Graham

Talk with Graham

tags: AI role-play / behavioural rubric

BRIEF

Graham is trying to send money to his partner overseas but the bank has blocked his payment. The situation is urgent as his partner needs the money soon.

You're the banker. Talk with Graham to understand what the payment is for, where it's going and notice the red flags as they crop up that indicate Graham is likely being scammed. But why would he believe you?

Earn Graham's trust. Protect his money. Escalate with care.

disclosure:

Graham doesn’t hand over the full story. He starts out worried and guarded, and opens up only as trust is earned. Ask about Daniel before the money, and he’ll tell you about the accident. Make it safe, and he’ll admit the doubts he hasn’t said out loud: the gift card request that didn’t sit right, and how every plan to meet falls through. Push too hard, too early, and he closes the door.

rubric:

Every call is scored on five behaviours: building rapport without shame, recognising the red flags, asking open questions, slowing the payment, and escalating with care. Learners are scored on what they do, not the words they choose, so different approaches can succeed. Releasing the payment, shaming Graham, dismissing him or making a discriminatory remark fails the call outright.

guardrails:

Graham never decides on his own that he’s being scammed. He can voice doubts, but reaching that conclusion takes a banker who listens, names what they’ve noticed, and offers him a way to check without feeling foolish.

This is how we design a character.

// our own process

How we use AI ourselves

  1. [01]

    Scope

    We talk with you and your subject matter experts, and agree what good looks like.

    human
  2. [02]

    Write

    The (sometimes unusual) ideas, characters, scripts, scenarios and questions are ours.

    human
  3. [03]

    Workshop

    AI pushes back on our draft, calling out gaps and unclear moments, and showing how different learners might read it.

    AI-assisted
  4. [04]

    Test

    AI checks across all formats, reading and providing feedback as different learners, picking up the ever elusive typo and checking builds to design plans. For AI role-play, AI plays the learner again and again (and again), so we find where the character breaks before your people do.

    AI-assisted
  5. [05]

    Refine

    We make your changes, checked against our SAFE Learning Design Framework and adult learning principles.

    human
  6. [06]

    Sign off

    You sign off the finished work.

    human

What AI never does here

  • Come up with the ideas
  • Write the scripts, characters or scenarios
  • Decide what scenarios are used
  • Make design decisions on sensitive content
  • Sign off on anything

Works with your tools

Our designs work within the AI tools your workplace already uses, and we work with you to keep your data protected.

Not using AI, or not allowed to?

We hold the creative process, and can adapt to your way of working.

Already using AI? We'll check what it made against our SAFE framework.

Or email aneeta@antelopecamel.com.au

Book a scoping call