Elite thoroughbred horse racing

The business goal

Our client breeds business elite racing horses with excellent physical characteristics. The goal of this project is to build an application that processes videos of horses and calculates the probability of being elite for the specific horse based on certain criteria. In order to do this we needed to prove the correlation between the available information about horses like cardio and biological information and their characteristics.

Among the main tasks were:

  • Build a web app that can utilize the developed predictive algorithms and manage ongoing data
  • Machine Learning module with high accuracy of prediction

Technical side

On a technical side we used Python and Django for web app and OpenCV with Neural networks for Machine Learning module.

The project consists of three basic parts:

1) Creating machine learning algorithms and models based on provided information about cardio, echo, DNA and eliteness information of the horses

2) Creating database and Web Application that manages current information and gathers new data about horses, predicts eliteness of custom horses and generates reports

3) Creating Web Jobs for automatic retraining of the machine learning models and rescoring them based on updated models and data from the database


As a result we delivered a web app that proves correlation with more than 80% of accuracy for elite thoroughbred horses based on different data like

  • cardio and biological information
  • heart rate, weight, body size, cardiac index , ultrasound/echo, DNA, etc.


Byron Rogers
Alex and his team did a great job completing what was a complex video/image recognition and machine learning task. They built a web app, database and ML pipeline that delivered exactly what I required with considerations to minimizing costs and ensuring that I had an outcome that adhered to goals. I would recommend their services.
Get in touch


    Contact us on Upwork

    Hire us

    Contact us on Linkedin

    Reach out us

    Moscow, 117335,