Work

One catalog, many storefronts

Product search, autosuggest, filters, and per-field similarity for an omnichannel retailer: web, marketplace, and stores.

The problem

The same goods had to be findable from a website, a marketplace, and the systems behind physical stores. One similarity function was not honest about every field.

The approach

I worked the relevancy of that backend, including similarity chosen for the field type, and moved the search platform across Solr versions and onto AWS.

RelevanceApache SolrEnterprise search

The conclusion

Omnichannel is a relevance problem before it is a channel problem. The field knows what 'close' means. The ranker has to follow.