A leading U.S.-based CrossFit coaching company, recognized for training elite athletes and professional coaches, faced significant challenges in scaling and streamlining its training plan creation process. Coaches relied on crafting personalized training plans in a freeform style using spreadsheets to maintain flexibility and speed. However, manually transferring these plans into the training platform was time-consuming, error-prone, and hindered the company’s ability to grow.

To overcome these hurdles, we built an AI-powered solution that integrates multiple cutting-edge technologies designed to improve efficiency and scalability. Our approach used OpenAI’s large language models (LLMs) and natural language processing (NLP) to convert freeform text training plans into structured data. This eliminated manual data entry, creating a fast, automated workflow that optimized coaches’ time.

To ensure the system could understand CrossFit-specific terminology and formats, we used Antlr to define a domain-specific language. This custom language allowed precise interpretation of coaches’ freeform input, turning natural language into data-ready formats with high accuracy.   Our solution included the Monaco Editor (built on the VSCode editor) to provide a user-friendly interface. With features like IntelliSense, syntax highlighting, and real-time error detection, coaches could easily create plans within a freeform environment while the platform automatically structured the data in the background.  For accelerated code development, we utilized Cursor to iterate faster on our AI and language processing models. Cursor significantly shortened the development cycle by enabling rapid feedback and testing, helping us refine the CrossFit language and improve accuracy.  To support continuous integration and deployment (CI/CD), we used Vertex for efficient pipeline automation, allowing the team to test, update, and deploy code seamlessly. This setup ensured consistent performance and minimized downtime as we iterated on the platform.

Implementing this AI-powered solution improved efficiency, reducing the time required to create and distribute training plans. The coaches experienced fewer errors and could focus more on coaching and athlete development. This scalable, AI-driven platform has set a new standard in athletic coaching, laying the groundwork for future enhancements like adaptive programming and predictive analytics to elevate athlete performance further.