OnCourtAI turns a short phone video into real biomechanics: body tracking, technique scores and coaching. Proven in tennis and padel, and ready for any sport or activity where movement matters. Build it into your own site or app.
// one video in → structured analysis out { "sport": "tennis", "stroke": "forehand", "shots": 4, "overall_score": 62, "components": { "preparation": 58, "contact_point": 71, "kinetic_chain": 64, "followthrough": 73 }, "coaching": ["Turn your shoulders earlier…"] }
One live product and three early-access integrations, all running on the analysis engine behind the OnCourtAI app. Early-access partners help shape what we build and get pilot terms.
An embeddable widget that recommends frames from your own catalogue, based on each customer's actual swing rather than a quiz.
Send us a video and get a full stroke analysis back as JSON: overall score, component scores, key moments and coaching tips, ready to show in your own product.
Plain computer-vision output for teams building their own datasets and models: body landmarks for every frame, joint angles and, for racket sports, shot segments.
Players upload a video on your club or coaching website. The analysis goes straight into the coach's OnCourtAI dashboard, ready for feedback, goals and drills.
On mantissport.com, players film a few shots on their phone. OnCourtAI analyses their technique and the widget, styled in Mantis's own branding, recommends the Mantis frame that suits how they actually play, with the reasons why.
See how it works →
Help online shoppers choose with confidence, and capture data on how your customers play.
Give members and squads video analysis between sessions, under your own name.
Collect player videos from your own website and review them in your OnCourtAI coach dashboard.
Add tennis and padel technique analysis to your app without building computer vision yourself.
Track technique across programmes and regions with consistent, objective measurements.
Build your own datasets from raw pose data, including for sports beyond tennis and padel.
Our computer vision tracks the whole body from an ordinary phone video, and that isn't limited to rackets. Some places where it could work for you:
A short call about what you want to build, which sports and strokes matter to you, and your expected volumes.
We issue API keys, set up your integration and run it on real footage from your players or customers.
Usage-based pricing agreed with you before launch, plus monitoring and support from the team that built the engine.
Every partner gets its own API key, restricted to the websites you specify, with usage and health monitored for you.
Results can be posted from our servers to yours, signed so your systems can confirm every result really came from OnCourtAI.
By default, uploaded videos are deleted straight after analysis. Retention is available per partner if you want to keep footage.
The same engine 950+ players have used in the OnCourtAI app, built and tuned stroke by stroke for tennis and padel.
The Stroke Analysis API, Raw Vision API and Club & Coach Embed are being built with our first partners. You'll work directly with us on a pilot using your own videos, help decide what goes in the first release, and get pilot terms. Racket Finder is fully live today.
Tennis (serve, forehand, backhand, volley) and padel (serve, forehand, backhand, volley, lob, bandeja, víbora, smash, chiquita). Racket Finder is tennis-only for now. For any other sport or activity, such as cricket, golf, pilates, physio or running, the Raw Vision API provides full-body tracking today, and we can build sport-specific scoring with you.
Not for Racket Finder or the Club & Coach Embed. Both are a single embed code we help you set up. The Stroke Analysis and Raw Vision APIs are for teams with a developer who can call an API.
A short clip on a normal phone, filmed from the side or behind, with the whole player in frame. No sensors or special cameras are needed.
Racket Finder has published monthly plans from £49.99. The early-access products are priced on usage (per analysis, or per video minute for Raw Vision), agreed with you before launch.
Tell us a little about you and we'll reply within one business day.
Thanks for getting in touch. We'll reply within one business day.
Prefer email? partner@oncourtai.co.uk
Start with Racket Finder today, or join the pilot for our new integrations.
Talk to us →