A personal insight.
A product direction.

Where it started

Ascelon’s founder began by using AI to analyse personal training data alongside wider context: cycling, running, strength training, recovery, weight and how the day felt. The experiment suggested that connected information could support more useful conversations about training.

A personal AI Coach followed. Ascelo grew from the wish to bring that kind of context into a product, without repeatedly assembling and pasting an athlete’s history.

The problem worth exploring

Athletes can have several goals at once: indoor cycling alongside strength work, improving cycling performance while gaining muscle, or balancing training around a long ride. A coach may only see the wider picture when the athlete explicitly explains it.

The opportunity: make relevant context easier to understand, while keeping athletes and coaches in control.

The platform and its apps

Ascelo is our fitness coaching platform and product brand. It brings together Ascelo Coach, a web app for coaches, and Ascelo Athlete, with separate iOS and Android apps for athletes.

Ascelo is in private beta by invitation. This is an early product journey, with feedback intended to shape development. The personal AI Coach is a separate founder-built project that informs the thinking; its implementation is not presented as a list of features available in the Ascelo beta.

Explore Ascelo Coach ↗ (opens in a new tab)   Visit the product website ↗ (opens in a new tab)

The AI direction

The immediate direction is coach assistance: synthesise permitted context and prepare guidance for a coach to review using their own knowledge and experience. Permission design and human judgement are central to that intention.

An independent-athlete AI adviser is a longer-term possibility. It is not a current release promise.

What this journey demonstrates

The work connects problem discovery, product design, architecture and incremental delivery. The aim is to build from a concrete need and test the value with people, rather than treat AI output alone as a finished product.