Increasing Client Retention by 52.3% with AI Coach for Strength Training

A fitness platform operating on USA Market (NDA client) struggled to retain high user engagement, prevent churn, and remain competitive in a tight market. The company decided to tackle this challenge by introducing an AI Coach that would provide personalized training programs to the users based on human pose estimation-based motion tracking.

They chose MobiDev to create and implement the Tech Strategy for the product development. After the launch of their HPE-based AI Coach, the company managed to increase its client retention by 52.3%.

Key Facts

USA

Country

Fitness

Industry

2025

Cooperation Period

AI Fitness Coach Development

AI Pose Estimation Development

Service Type

The Story Behind

Our client, a USA-based Fitness Application for Strength Training, was launched during pandemics. For a long time, it served as a platform for people who wanted to continue their exercises but couldn’t attend the gym due to the lockdown. After the restrictions were lifted, the company had been facing churn issues as it struggled to deliver additional value for the customers with standard training plans.

The company decided to offer its client base a more personalized approach with a custom AI Fitness Coach. They searched for a tech development partner with extended expertise in Multimodal LLMs and Human Pose Recognition. They chose MobiDev due to our proven experience in building AI and HPE apps for fitness and sports.

Business value

After COVID-19, the fitness market has become saturated with applications, and it’s becoming increasingly difficult to stand out. Despite a steady demand for apps that provide exercise guidance wherever and whenever users want to train, most applications experience churn once a user has “seen and tried everything there is to see and try.” At the same time, the expectations for deeper personalization are growing, bolstered by digital experiences in other industries, primarily retail.

Project Scope

The project goal was to increase client retention by introducing an AI coach that would use Human Pose Estimation for motion detection and analytics and the available database of training programs to offer personalized training experiences for users.

The project began with AI Consulting, which included an app audit and Tech Strategy. Next, we allocated a dedicated team of 6 technical experts to implement the strategy. We chose the Human Pose Estimation model that best met the project goals. Next, we trained Pose and Movement Analysis model for HPE data interpretation.

We also created a coaching logic using outputs of the Pose and Movement Analysis Model with event-driven approach to provide real-time feedback and instructions. The Project lasted 12 months, with the MVP ready in 6 months after the beginning.