Can an AI coach feel like a living system instead of another dashboard?
I’m building APEX PULSE PRO around a question I keep coming back to:
What if an AI coach didn’t just generate workouts, but visibly changed with the person using it?
The center of APEX is something we call the Core.
It already reacts to the athlete’s inputs, recent training activity and interaction state. Instead of treating it as a decorative AI orb, I want to push it further - into something that feels more like a living performance system.
The direction I’m exploring now is a neural layer that grows visually around the Core.
Not an avatar replacing the AI.
Not a medical visualization.
Not another dashboard full of scores.
The Core remains the intelligence.
The neural layer would simply make its internal state easier to feel - warmer and more compressed when the system is under load, gradually moving toward more energetic/cooler expression as the athlete’s APEX state improves.
I’m also experimenting with a concept called APEX Position: a simple way to show where the system currently places you, why, and whether your direction is improving - without pretending it’s a medical readiness measurement.
I’d really value the perspective of other makers here:
Would this make an AI coach feel more understandable and motivating - or does it risk becoming visual complexity that gets in the way of the product?
If anyone wants to try APEX and give product feedback, I’m happy to share it in the discussion.
Replies
The instinct is right, state should be felt before it's read, but ambient color and compression only work if people already know what warm-and-compressed means for them. That's a private language you have to teach, and coaching products live or die on how fast someone trusts the read. I'd keep the Core as the feel layer and put one plain number or word next to it, so the visual builds intuition over time instead of asking for it on day one. APEX Position could be that anchor.