Forum Discussion
Powerful Positioning with AI
Built as a high energy portfolio concept, this project is an interactive Storyline experience that helps learners practice “Power Positioning” through short scenarios, branching choices, and coaching style feedback inside the module.
Behind the scenes, I approached it like a real client build: clarified the performance need and audience, mapped the learner flow, drafted a lightweight storyboard, prototyped quickly, then iterated based on usability checks to keep the experience fast, clear, and visually consistent.
What makes it stand out is the AI integration. The experience uses a ChatGPT API to analyze the learner’s typed response, generate custom strengths and next steps, and provide a 1 to 10 score, giving learners targeted coaching they can immediately apply on the job.
A Sample Responses Key is available here.
At a high level, the Storyline setup is pretty simple: I use a text entry field tied to a variable, a JavaScript trigger on a button to send that text to an API endpoint, and then a few Storyline variables to capture the response and drive the feedback layers.
I’ve pasted my Storyline setup and the JS trigger into the attached txt file. Your team could point the same pattern at your internal Copilot endpoint instead of ChatGPT.
Storyline 360 setup
- Variables in Storyline
- UserInput – Text (bound to the text entry field where the learner types their recap)
- Score – Number (1–10 from the AI)
- Band – Text ("low", "mid", or "high" for branching)
- StrengthFeedback – Text
- OpportunityFeedback – Text
- NextStepFeedback – Text
- (Is_Loading – optional True/False variable for a loading layer)
- I created these Storyline variables:
- Slide elements
- A Text Entry field bound to UserInput
- A button labeled something like Get AI Feedback
- A Loading layer that shows while the call is happening
- Three feedback layers (Low / Mid / High) that read:
- Score, Band
- StrengthFeedback, OpportunityFeedback, NextStepFeedback
- Branching logic
- If Score <= 3 → show Low Feedback layer
- If Score between 4 and 7 → show Mid Feedback layer
- If Score >= 8 → show High Feedback layer
- On the Loading layer I use triggers like:
- What the JavaScript does
- It reads UserInput
- Sends it to a single API endpoint as JSON
- Receives JSON back with a score and three coaching messages
- Writes those values into the Storyline variables listed above
- On click of the Get AI Feedback button:
Your IT team could point that same pattern at your internal Copilot agent as long as they expose an endpoint that accepts the learner text and returns JSON.
- Variables in Storyline
14 Replies
Very cool!
This is very cool TrenWalker! The AI-driven feedback is especially interesting here. It's neat how it's using responses to give next steps and a score. I can see this being repurposed for various scenarios, such as training a new hire and providing feedback.
We’ll be featuring this in the ELH Weekly newsletter. If you want it delivered to your inbox, be sure you’re subscribed. 🎉
Thanks so much!
Hi Tren - this is fantastic! We have an internal Copilot agent that we have restricted to use particular sources that would work perfectly with this idea. I'm interested to learn how you integrated the prompt for Storyline to use ChatGPT. I'll work with my IT team on an API but I'll need to do everything from the Storyline end so would really appreciate more information on how you did it.
Thanks so much, I'm glad you liked it!
On the Storyline side it's pretty simple: the learner types their recap into a text field, and a JavaScript trigger sends that text to a single API endpoint. That endpoint (for me, a small Cloudflare Worker) calls ChatGPT, then sends back a score and coaching points that I drop into Storyline variables and show on different layers.
The API key itself lives securely on the Cloudflare side as a secret - Storyline only ever talks to the endpoint, not the key
Happy to share a quick overview of the Storyline setup and JS trigger if that would be helpful for you and your team.
That's good news for me :) Yes, please share the setup and JS trigger - I really appreciate you sharing.
You are an absolute legend!!!! I appreciate this SO much, I think this could be a game changer in our organisation. There was already a lot of excitement over the Agent, this will take it to the next level. I'll let you know how we go. Unfortunately, I won't be able to share the finished product as I work for a pharmaceutical company and would be breaking the law, but I can certainly let you know how it's received.
This absolutely made my day, thank you. I am so excited to hear there is already buzz around your Agent!
I totally understand about not being able to share the finished product. I would just be thrilled to hear how it lands and what your teams think of it.
Selfishly, I have also been experimenting with integrating AI like this as a portfolio boost while I look for my next role, so it means a lot to hear that it resonates with someone in the field.
Please do let me know how it goes, and congrats in advance on whatever you build with it.
I'm so glad this has made you happy 😁 Not at all selfish to build yourself a niche, definitely the way to move with the pace of change. I suggest you have a play with Agents if you're not already, we have developed a couple of Copilot ones that are really helping in the field.
I haven’t used copilot, I’ll have to check it out.
Really nice implementation, TrenWalker The Cloudflare Worker approach to keeping API keys secure is spot-on - I've seen too many people make the mistake of exposing keys.
I've been working on something adjacent - voice-based AI roleplay that embeds into Storyline. The learner speaks, the AI responds with voice, and the whole thing evaluates against learning objectives. Different modality, similar philosophy.
Would love to compare notes if you're interested!
Here's a quick demo of how it works.Wow...this is absolutely phenomenal! I’m really inspired by the work you’ve done here. I’ve started creating a practice project file to learn from, and I’d love to build something like this for my own portfolio. If you’re open to it, I would really appreciate the opportunity to connect and learn more about your process. Mentorship on this would mean a lot!
I’m happy to help. You should be able to find me on LinkedIn.
