Mo, a running partner that helps you choose your next shoe
There are hundreds of running shoes, but most shopping tools still answer with a long list of results. Runners end up comparing specs, prices and reviews on their own. Mo is an independent AI running partner. A store knows its catalog, but Mo knows the runner.
Each one explains why it fits your rotation, with the price and where it belongs.
The problem
Runners don't need more search results. They need to know which shoe to buy next, and whether they need one at all.
The shift
An agent, not an advisor. Mo speaks up first, checks the rotation, narrows it down, and points to a store.
The result
A live prototype where the model chooses three picks and explains each one.
Runners don't need more results, they need a decision
Most runners keep a rotation: one shoe for easy runs, one for speed workouts and one for race day. So the real question isn't only which shoe is good, it's which shoe they need next.
Today that decision falls on the runner. Lots of options, confusing specs and not knowing when to replace a shoe make it easy to buy the wrong thing, or to put off buying anything.
Mo is for the intermediate runner training for a half or full marathon, with two or three shoes in rotation, at the moment they want something new: their usual model changed or sold out, or they're curious about another brand.
The best shoe buying help is a good human in a running shop
Mo is that, without the shop. It isn't owned by a store, so it's on the runner's side. I summed up the positioning as taste plus safety: taste gets the runner to open it, safety is why they trust the pick.
Check the rotation before recommending anything
Recommend a good shoe, even if it does the same job as one the runner already owns.
Starts with the shoes the runner owns, notices a missing shoe type, and says so once. If there's no gap, Mo says that too.
Three picks, not a list
Dozens of results, and the runner still has to do the comparing.
One top pick and two alternates, each with the price, what it's for, and why it fits this runner's rotation.
Ask in three taps instead of connecting Strava
Connect Strava or Garmin to read mileage automatically.
The runner adds their shoes in three taps and a rough mileage once. Mo estimates forward and says "probably" when it guesses.
A live prototype with a real model behind the picks
Model-chosen picks
Claude Haiku 4.5 chooses the picks and writes the reason on each card. Scripted picks step in if the call fails.
Remembers the runner
The rotation is saved in the browser. Mo opens with "Welcome back" and names the shoe that's probably due.
Around the run
A race question with a countdown, and a store finder by ZIP code. Mo can't see stock, so it says to call ahead.
Test the recommendations with intermediate runners, add more tempo shoes to the catalog, and let runners add a shoe from a photo. After shoes, Mo can grow into training plans, tips, and gear beyond shoes.