# How Amazon Uses LLMs and COSMO to Improve Product Recommendations

DevFeed: [How Amazon Uses LLMs and COSMO to Improve Product Recommendations](<https://devfeed.tech/articles/how-amazon-uses-llm-for-product-recommendation-18124.md>)

Original publisher: [Read original article](<https://hungrymindsdev.substack.com/p/how-amazon-uses-llm-for-product-recommendation>)

Author: Alexandre Zajac

Published: 2026-05-04T15:30:20Z

Content type: article

Language: en

Sources: [Hungry Minds](<https://devfeed.tech/sources/hungry-minds.md>)

Topics: [amazon](<https://devfeed.tech/topics/amazon.md>), [recommendation systems](<https://devfeed.tech/topics/recommendation-systems.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [data](<https://devfeed.tech/topics/data.md>)

Tags: [amazon](<https://devfeed.tech/tags/amazon.md>), [data](<https://devfeed.tech/tags/data.md>), [knowledge-graphs](<https://devfeed.tech/tags/knowledge-graphs.md>), [llm](<https://devfeed.tech/tags/llm.md>), [products](<https://devfeed.tech/tags/products.md>), [recommendation-systems](<https://devfeed.tech/tags/recommendation-systems.md>)

## AI overview

The article explains how Amazon built COSMO, a commonsense knowledge graph for product recommendations. It describes using an LLM to generate candidate explanations from shopping behavior, filtering and evaluating those candidates, deriving a structured ontology, and distilling the result into COSMO-LM.

## Source excerpt

PLUS: How Shazam Works 🎼 , Agent Harness 🧇, Vercel Breach 🔥