Case Study: Kwiat
Helping Kwiat improve product discovery and delivering a better search experience
We had a chance to sit down with Cory Kwiat, the Chief Digital Officer for Kwiat for a chat. For over a century, the Kwiat family has been a fixture in the jewelery industry with their exquisite design and craftsmanship. "We own and operate two different high-end luxury jewelry brands, one being Kwiat Diamonds and the other is Fred Leighton, which is a vintage and estate collection. We also have retail stores here in the US - one in New York and one in Las Vegas. We have a wholesale distribution and manage online retail as well," says Cory.
Cory gives some background about the tech-stack the family business was using, "We had been using Access which is pretty antiquated, but we had built a customized system on top of it to help run the business. We were supporting it for a while, but later on, we migrated to an ERP system integration which handles everything from accounting, inventory management, production management. "Our ERP system is not user-friendly and built for the type of searching we needed. This was causing problems for our users who were trying to find inventory items. The big challenge this has in our industry is that, in jewelry everything is very visual. So, my goal was to develop a technology that where we could store data in our ERP system while giving users a search experience that they, and we require. I wanted to use Elasticsearch as I had some previous projects, but needed some dependence on a developer or coder for certain needs but this project didn't afford me the support."
"Once we had our ERP in place," Cory continues, "we had originally developed a whole interface to do what we needed but since it was being done by a third party, we never got to perform just right. It was too slow and too limited. So, I thought of dual tracking a backup plan and think about the problem differently. Instead of using the existing framework and datastore, we can work on exporting that data out of our internal ERP and pushing that into Elasticsearch and bridge the two systems. "So, this freedom allowed us to look at some solutions and that's where ReactiveSearch.io came in. When I found ReactiveSearch.io, it seemed like the perfect Elasticsearch as a service that we are looking for Kwiat. It checked a lot of the technological requirements that one would experience with Elasticsearch and took them out of the equation and allowed someone like me, while I'm technical, I'm not a coder, to be able to think through and develop the whole framework and architecture that we utilize. Once we got through the initial R&D, development and changes were quick. We have a complex search in place which is nuanced in its logic and filtering, but nothing we couldn't handle after putting in a bit of time and effort."
"ReactiveSearch.io seemed like the perfect Elasticsearch as a service that we are looking for Kwait. It checked a lot of the technological requirements that one would experience with Elasticsearch and took them out of the equation."
Cory KwiatChief Digital Officer, Kwiat
"Using ReactiveSearch, we were able to add customization on top of our integration and solve what we viewed as a complicated problem quickly and cost affordably. In just a few hours, I was able to start on a proof of concept and show the rest of my team what we wanted to do and how easy it should be. The tutorials were a great help in showing how simple the process can be." He continues, "On the subject, I also wanted to point out that the support has been incredible and the responsiveness was instrumental in giving us the confidence to get to production ready. Previous projects where I've helped develop the search would take weeks if not longer in just building the tech stack and understanding the storage and that whole process was reduced to just a few days.
"With the frontend being handled by React, we had to send our data to Elasticsearch. Thanks to the ABC-CLI tools, we were able to think through how to connect our internal system to Elasticsearch. We are using a chron job that checks every 5 mins for changes to our index and sends updates to the system," says Cory.
Like I mentioned, ReactiveSearch.io takes what otherwise would require a real expert in Elasticsearch and administration out but doesn't take the control or access away from someone which I think is great. That works for someone who is not as tech-savvy and someone who is. Moving forward, we're going to refine the data integration by building a system that will push data automatically to Elasticsearch. We're also planning to release an iOS app in the next year and looking into Role-Based Access Control and SSO will be vital since it will most likely require its own authentication framework," Cory concludes. ReactiveSearch.io features that Kwiat leverages today:
- ReactiveSearch - UI components for building search
- ReactiveSearch.io's Access Control settings to create a secure search experience
- Search Relevancy Control Panel to visually build and test relevance settings
- ReactiveSearch.io's ABC CLI tools to import data from an extensive number of databases
"In previous projects, the search would take weeks if not longer in just building the tech stack. That whole process was reduced to just a few days."