Badminton coaching assistant

Welcome to the Badminton Coaching Assistant! This application is here to inspire and help you improve your badminton gameplay by teaching you techniques and strategies to achieve your full potential.

Comment

Inspiration

My mission is to enrich the experience and elevate the proficiency of badminton players. I aim to help players overcome the typical hurdles that they face, such as lack of motivation or difficulty executing certain techniques. Additionally, I aim to inspire and encourage players who are struggling and provide them with the support and resources they need to reach their full potential on the court.

What it does

The Badminton Coaching Assistant is an exceptional tool that offers extensive support to badminton enthusiasts striving to achieve their full potential. With advanced techniques and strategies, this assistant enables players to elevate their gameplay to the next level. Its comprehensive guidance and personalized coaching make it an ideal solution for both novice and seasoned players.

How we built it

I utilized the cutting-edge technology of PartyRock, which is powered by Amazon Bedrock, to develop this application. I have employed the highly effective Claude family foundational model to ensure the app's superior performance and reliability.

The text prompts were optimized and fine-tuned according to AWS Bedrock Prompt engineering guidelines.

The user interface for the question-answer system is designed with three sections. The first section allows the user to input their question. In the second section, the system generates a response for the user. This response is presented in two formats - a detailed long-form answer and a concise summary. The third section enables the user to review and evaluate the system's response in both formats, thereby facilitating a more informed decision-making process.

[link] /badminton-coach-architecture.png

Challenges we ran into

During my research, I have been conducting experiments with different models, namely the Claude, Titan, and Llama models. To ensure that I get the best possible results, prompt engineering has also been crucial in fine-tuning the responses of these models. By combining both experimentation and prompt engineering, I aim to enhance the overall performance of these models and provide better results.

Accomplishments that we're proud of

Build a Generative AI application to help badminton players improve their skills quickly. Learn Generative AI in the context of the AWS tech stack. Use Prompt Engineering Techniques.

What we learned

By harnessing the power of AWS tech stack and using Prompt Engineering techniques, one can create an advanced application that can elevate the proficiency of badminton players to new heights.

What's next for Badminton coaching assistant

I would like to enhance the functionality of this application by integrating AWS Rekognition Service. This will enable players to benefit from AI-powered game play analysis, such as providing statistics on the number of hits, hit positions, number of out-of-bounds hits, and number of smashes. Additionally, the service will provide qualitative analysis of player performance during the game.

The other aspect of extending this application is integrating Amazon Shopping with recommended products associated with user questions and responses from the model.

Built With

Share this project:

Updates