AI in the browser: predicting user actions in real time with TensorflowJS [eng]

With AI becoming increasingly present in our everyday lives, the latest advancements in the field now make it easier than ever to integrate it into our software projects. In this session, we’ll explore how machine learning models can be embedded directly into front-end applications. We'll walk through practical examples, including running basic models such as linear regression and random forest classifiers, all within the browser environment.

Once we grasp the fundamentals of running ML models on the client side, we’ll dive into real-world use cases for web applications—ranging from real-time data classification and interpolation to object tracking in the browser. We'll also introduce a novel approach: dynamically optimizing web applications by predicting user behavior in real time using a machine learning model. This opens the door to smarter, more adaptive user experiences and can significantly improve both performance and engagement.

In addition to the technical insights, we’ll also touch on best practices, potential challenges, and the tools that make browser-based machine learning development more accessible. Whether you're a developer looking to experiment with ML or someone aiming to bring more intelligence into your web apps, this session will offer practical takeaways and inspiration for your next project.