Performance intelligence · 01
Padel Insights
Computer vision for players, coaches and clubs.
See the shot. Read the body. Build the player.
The opportunity
Coaches and clubs are stretched thin, and players get only minutes of hands-on feedback per session. Players can't see their own technique objectively, and coaches can't track movement, workload or injury risk once the player walks off court.
Padel Insights changes that: it turns every match and practice into computer-vision stroke analysis, flags injury risk before it becomes pain, and builds a personalised training program — complete with targeted exercises — so players improve faster and stay on court longer.
The solution
Pose-estimation overlays on volleys, kick smashes, smashes and split-step timing — frame-by-frame or side-by-side with reference technique.
Plain-language notes on what the footage shows and why: “late split-step reduces reach on wide balls.” Coaches review, edit or endorse.
Structured drills and focus areas generated from footage, which a coach adopts as-is or adjusts to fit their own programme.
Mobility, strength and recovery suggestions built from movement-pattern data — a supplement to the physio or S&C already in place.
Surfaces overload and compensation patterns — shoulder and elbow strain indicators — early enough for coaches and physios to act.
Heatmaps, rally length, shot selection and technique trends over time: an evidence base behind every player's plan.
Built with coaches
Every insight is drafted for a coach to review, edit or endorse before the player sees it. The relationship stays theirs; the evidence base gets stronger.
How it works
Padel Insights is built around three connected layers: it sees what a coach would miss, spots the strain before it becomes an injury, and turns that into a program that is yours — not a generic template.
Phone or court footage is converted into a skeletal tracking breakdown of your strokes, footwork and court position. We measure split-step timing, contact height, joint angles and movement patterns — so you can see exactly what is happening in every phase of a shot, not just guess from the result.
The same movement data that improves technique also reveals overload. We flag compensation patterns, shoulder and elbow strain, and sharp workload spikes before they turn into pain. Coaches and physios get an early warning, plus targeted mobility, strength and recovery exercises to rebalance the load.
Insights become a weekly plan: on-court drills, off-court exercises, and focus areas chosen from your own footage. The coach reviews, edits or endorses every recommendation — so the program stays personal to your game, your body and your goals, not a one-size-fits-all checklist.
Worked example · Kick smash
Here is what the engine does with a single shot: track the movement, explain the fault in plain language, then hand back the drills and exercises that fix it.
Contact height
2.42m
above ideal window
Forward jump
1.1m
into court
Racket-head speed
High
top 15%
Landing load
3.2x BW
front leg dominant
This is an illustrative AI-generated clip — the engine tracked take-off, contact and landing, then wrote the metrics and steps on the right. Your clip gets the same treatment: upload one short clip of your shot and it returns your metrics, three improvement steps with drills, and any injury risk worth watching.
MP4 or MOV, up to 30MB — one shot is enough. Results in under a minute.
The skeleton tracks a jump that lands 1.1 metres inside the court, contact 2.42 metres above the ground and racket-head speed in the top bracket. The front leg takes 3.2x bodyweight on landing, with the back leg trailing.
The forward jump converts height into depth: the ball is hit above the ideal contact window and driven down toward the back glass, making it hard for opponents to recover from deep behind the baseline.
Cue the take-off from the back leg so the body travels forward, not just up. Reach the racket high and behind the head early, then snap over the ball at the peak of the jump. Land on the front foot ready to volley the reply.
3 sessions this week: approach-footwork reps (10 min), jump-and-contact height drills on a lob feed x25 (12 min), plus a single-leg landing and calf-hamstring strength circuit to manage the flagged landing load.
Try it · Upload a clip
Drop in a short clip of your own padel — a kick smash, a vibora, a serve — and the engine returns what it sees, the key metrics, three steps to improve with drills, and any injury risk worth watching.
Want it tailored to you? Fill in your player profile — your height, weight, age and playing style shape the metrics, drills and injury advice, and you can pick a favourite pro (Arturo Coello, Agustín Tapia and more) to model your technique on.
Your improvement steps appear here
The engine reads your movement, scores the key metrics, writes three improvement steps with drills, and flags any injury risk it spots.
In the app
Players record, review and follow the plan. Coaches endorse the advice and watch the whole roster's workload from the same data.
Split
-0.14s
Contact
2.42m
Load
High
What the footage shows
Take-off is slightly late, so the jump is more vertical than forward. Racket arm reaches the ball at 2.42m on 6 of 8 kick smashes.
Reviewed by Coach Marta ✓
Record any clip and get frame-by-frame skeletal tracking.
Business model
Ad-supported free tier for self-analysis. Premium removes ads, unlocks full plans and lets players share footage and progress directly with their coach.
Individual players, with or without a coach
Automated recording plus a coach dashboard to review skeletal-tracking data, edit AI-drafted plans and monitor a whole roster's workload.
Padel clubs, academies & coaches
A rebrandable analysis engine licensed to court-camera vendors, booking platforms, federations and coaching apps as their movement-analysis layer.
Hardware vendors, federations, booking apps
Why now
Go-to-market
Launch with 10–15 pilot clubs and their resident coaches across the UK and Spain, positioned explicitly as a coach's assistant. Coach buy-in seeds adoption of the D2C app among their players, while OEM conversations run in parallel with camera and booking-platform vendors who need an AI layer to differentiate.