Future-proof web apps with Angular and TensorFlow.js

on YouTube ↗14m

" summary

Shows concrete glue: Angular templates use TensorFlow.js models via an npm/CLI integration that exports a pipe to run inference off the main thread so a toxicity classifier can emojify input without dropping frames. A production pipeline trains site‑specific prefetch models from Google Analytics → BigQuery → Dataflow → TFX and exports TFJS models that lazy‑load, support quantization (up to 4× size reduction) and power the predictive prefetching demo that targets measurable latency and conversion gains.

AI-written summary · about

" chapters (5)

1. 0:00Start
2. 4:57implement the entire app by using angular
3. 10:19create a tfgs model
4. 12:47integrate our tfgs model with angular
5. 13:20build a custom machine learning model

" related talks

• 26mMachine Learning and AI for Angular DevelopersJerry Kurata · NG-MY 2019
• 24mWeb Apps of the Future With Web AIJason Mayes · JSNation 2024
• 23mData science with angularAri Lerner · ng-europe 2016
• 20mSmart Angular Elements An ExperimentCarmen Popoviciu · ngAtlanta Conf 2018
• 23mNode.js = Machine LearningNikhila Ravi · NodeConf EU 2017
• 18mMad science with the Angular CompilerMinko Gechev · ng-conf 2017
• 32mData science in the browser: DX & UXJana Beck · DinosaurJS 2019
• 22mBuilding Fast Angular Applications from EndMinko Gechev · JSNation 2019