Project Story About the Project The idea for this project stemmed from the desire to enhance the MLB fan experience using cutting-edge AI and cloud technologies. Traditional sports experiences often lack personalization and real-time insights, and this project seeks to address those gaps. By leveraging Google Cloud's robust machine learning tools, combined with real data from MLB, we aim to transform how fans engage with the game, making the experience more dynamic and interactive.
Inspiration As a fan of both technology and sports, the inspiration for this project came from realizing the untapped potential of AI in personalizing sports content. Baseball, like many sports, has a massive fan base, and many of them crave more tailored interactions—whether that be personalized highlights, real-time commentary, or even in-depth stats analysis. Google Cloud's suite of tools like Gemini, Vertex AI, and Imagen made it possible to build a scalable solution that could bring these ideas to life.
How I Built It The project was developed using a combination of the following tools:
Google Cloud AI: Used for building predictive models, video analysis, and generating real-time insights. ML models: Custom-trained models for prospect prediction and real-time baseball strategies. MLB Data: The project utilized official MLB datasets to gather player stats, historical game data, and more to drive insights. Tech Stack: Python for data processing, TensorFlow and PyTorch for machine learning, Google Cloud Platform for deployment. The building process involved integrating multiple datasets and developing machine learning models that could predict outcomes, generate video highlights, and provide insights for viewers. I also ensured that the application was user-friendly by adding a simple, responsive interface for users to interact with.
Challenges Faced The biggest challenge was ensuring the real-time application could process and display data efficiently without delays, as performance is crucial in a live sports context. I also had to overcome challenges related to data preprocessing, as MLB data is massive and requires careful handling. Building models that could generate relevant predictions and insights was another hurdle, especially ensuring they worked well with the vast amount of historical and real-time data.
Log in or sign up for Devpost to join the conversation.