Eliran Natan
Build UIs that learn - Discover the powerful combination of UI and AI
#1about 2 minutes
Solving slow load times in single-page apps
React's lazy loading API helps with initial load times but introduces delays when components are actually rendered.
#2about 3 minutes
Prefetching components by predicting user actions
Improve user experience by using dynamic imports to prefetch components or entire routes before the user requests them.
#3about 2 minutes
Applying machine learning to user sequence prediction
Frame user navigation as a sequence prediction problem where the model learns from click history to predict the next action.
#4about 4 minutes
Implementing a neural network with TensorFlow.js
Build a neural network in the browser using TensorFlow.js with LSTM layers and one-hot encoding to handle sequential user interaction data.
#5about 1 minute
Training the prediction model asynchronously in the browser
Use the asynchronous `.fit()` command in TensorFlow.js to train the model with user data in the background without blocking the UI.
#6about 7 minutes
Architecting the solution with the React Context API
Use a `PredictorContext` for ML logic and a `PrefetchContext` as a message bus to decouple prediction triggers from prefetching actions across the application.
#7about 3 minutes
Summary of key techniques for a self-learning UI
A recap of combining React's code splitting, TensorFlow.js for prediction, and Context API for a robust, performance-oriented architecture.
Related jobs
Jobs that call for the skills explored in this talk.
Matching moments
48:43 MIN
Implementing code splitting and lazy loading for faster initial loads
How Your Bundle Size Affects The Climate
28:28 MIN
Demo code and using web workers for performance
Privacy-first in-browser Generative AI web apps: offline-ready, future-proof, standards-based
08:01 MIN
Why developers and businesses choose to use React
React and the power of visualisation
02:25 MIN
Solving UI challenges with a modern tech stack
Building Interactive Async UI with React 19 and Ariakit
25:00 MIN
Best practices and the future of browser AI
Privacy-first in-browser Generative AI web apps: offline-ready, future-proof, standards-based
24:03 MIN
Enhancing runtime performance with CSS and UX
The Illusion of a Performant Web Application
03:03 MIN
Building with familiar React APIs and real CSS
Lynx: Native for More
13:51 MIN
The technology behind in-browser AI execution
Generative AI power on the web: making web apps smarter with WebGPU and WebNN
Featured Partners
Related Videos
Catching up on the basics you don't really need that much code
Chris Heilmann
Building Interactive Async UI with React 19 and Ariakit
Aurora Scharff
The Journey of a Pixel in a React Application
Shem Magnezi
Snappy UI needs no Single-Page Application
Clemens Helm
Accessibility in React Application
Julia Undeutsch
Multiple Ships to the Island - Micro Frontends & Island Architectures
Florian Rappl
From ML to LLM: On-device AI in the Browser
Nico Martin
A journey of a long list in React
Adam Klein
Related Articles
View all articles
.webp?w=240&auto=compress,format)


From learning to earning
Jobs that call for the skills explored in this talk.
Frontend Web Developer (JavaScript, React)
Die Socialisten
Vienna, Austria
Senior
Jest
React
Redux
JavaScript

Staff React Engineer (TypeScript + AI Experience)
Wizeline
Senior
DevOps
Next.js
GraphQL
Terraform
TypeScript
![Fullstack Engineer [React.js, TypeScript, Inline Styling]](https://wearedevelopers-staging.imgix.net/staging/public/default-job-listing-cover.png?w=400&ar=3.55&fit=crop&crop=entropy&auto=compress,format)
Fullstack Engineer [React.js, TypeScript, Inline Styling]
CrowdBuilding
DevOps
Node.js
TypeScript
React Native
Microservices
+1



Web Entwickler:in Next.js / React
NUUK GMBH
Hamburg, Germany
GIT
Next.js
Terraform
TypeScript
Continuous Integration



Frontend dev react.js / Next.js Blockchain en AI
Multiplied
The Hague, Netherlands
Remote
Intermediate
HTML
Next.js
Blockchain
TypeScript
+1