marqo

Why we included this project

Marqo is a semantic and tensor search engine you can point at text, images, or mixed media and query in plain language. That makes it a natural fit for teams building product search or discovery features: its tensor approach blends dense embeddings with the familiar ranked-results model of a classic search index, so relevance stays flexible instead of being locked to keyword matches. The project grew out of ecommerce work in fashion, beauty, and other retail catalogs, and its examples show how to set up image search and personalized ranking against a product catalog, which helps when you are prototyping something similar. One caveat before you build on it: the open source repository is deprecated and no longer receives updates, so treat it as a reference and a starting point rather than a dependency for a long-lived production system.