A comprehensive guide on building a TikTok-like real-time personalized recommender system, detailing the architecture, including the 4-stage recommender model, and the two-tower neural network design. It uses an H&M retail dataset for practical application, teaches feature engineering, model training, and serving using the
Table of contents
A quick introduction to the H&M retail datasetCore paradigms for personalized recommendationsIntroducing the two-tower embedding modelUnderstanding the 4-stage recommender architectureApplying the 4-stage architecture to our H&M use casePresenting the feature/training/inference (FTI) architectureApplying the FTI architecture to our retail use caseDeploying the offline ML pipelines using GitHub ActionsQuick demo of the H&M real-time personalized recommender3 Comments
Sort: