Walmart Patent Granted for ML Base-Variant Grouping

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published March 24th, 2026

Detected March 25th, 2026

Summary

The USPTO has granted Walmart Apollo, LLC a patent for a scalable pipeline for machine learning-based base-variant grouping. The patent, titled 'Scalable pipeline for machine learning-based base-variant grouping,' was granted on March 24, 2026, and details a system for grouping candidate items using machine learning techniques.

View original document View source feed page

What changed

The United States Patent and Trademark Office (USPTO) has granted Walmart Apollo, LLC a patent (US12586347B2) for a novel system and method related to machine learning-based base-variant grouping. The patent describes a scalable pipeline that utilizes Siamese models, adjacency lists, graph generation, and hierarchy dendrograms to identify and group variant items based on density and distance thresholds. This technology aims to improve the efficiency and accuracy of clustering and grouping large datasets of candidate items.

This patent grant is primarily an intellectual property development for Walmart and does not impose direct compliance obligations on other entities. However, companies operating in the e-commerce, retail, or technology sectors that utilize machine learning for product categorization, recommendation systems, or data analysis may find the disclosed methodologies of interest for potential competitive benchmarking or future technology development. No immediate compliance actions are required for external parties.

Source document (simplified)

← USPTO Patent Grants

Scalable pipeline for machine learning-based base-variant grouping

Grant US12586347B2 Kind: B2 Mar 24, 2026

Assignee

Walmart Apollo, LLC

Inventors

Yanxin Pan, Swagata Chakraborty, Ekaterina Pirogova

Abstract

A system comprises one or more processors and non-transitory computer-readable media storing computing instructions that, when executed, perform operations comprising: generating an adjacency list for candidate items using a distance threshold with the maximum allowable neighbor distance equal to a Siamese model cut-off. The operations can also comprise loading data for the candidate items in the adjacency list and generating graphs of the candidate items in the adjacency list. The operations further can comprise determining, using breakdown logic, first graphs of the graphs that exceed a predetermined size, and building hierarchy dendrograms of nested subclusters of the first graphs. The operations additional can comprise determining cut-off values based on p-th percentiles of density for the first graphs, and identifying recommended variant groups of the candidate item in the nested subclusters of the hierarchy dendrograms below the cut-off values. Other embodiments are disclosed.