The problem

The search experience on e-commerce sites are stuck in the past. They utilize basic text matching and category matches but ignores all of the product metadata and user insights that e-commerce sites have access to.

The solution

We built the autocomplete brain and you plug into us. Our team has in the past, built a similar autocomplete system for a hotel search company, which boosted conversion tremendously. We realized that it is a feature that many e-commerce companies do not have the time to invest in. Our autocomplete service takes all of the product metadata and filters available from your e-commerce site and intelligently chooses the ones that work best for your shopper's queries.

What we built at SF Disrupt 2015

A proof of concept of what Quarry could do using the Walmart Open API for an initial data set. We connected our backend autocomplete API to a Google Chrome extension, so we could demo our autocomplete on top of

Here is a side-by-side comparison of our autocomplete vs the existing autocomplete on

Autocomplete plugged into Quarry's API


Original search on


The links on our autocomplete are linked to directly to with the correct filters applied, whereas the existing autocomplete mainly uses a text comparison to grab the top results.

Future goals

  • Generalize how we store data so we can partner with more sites
  • Integrate user analytics to surface the most popular searches
  • Enhance autocomplete matching algorithms
  • Build pipeline to process analytics data and update match scores
  • Leverage a larger set of shopping query data to further improve suggestions
Share this project: