KRS.io Builds Embedded Analytics Solution with Amazon QuickSight

KRS was held back by its legacy databases and legacy Hadoop clusters that didn’t scale. Hosting physical servers in an Equinix datacenter restricted growth. In order to achieve the CEO’s mandate to build the most advanced retail analytics platform in their space, KRS embarked on a two year plan to fully migrate their data center to AWS. As the first step, KRS’s Engineering team was tasked with migrating 30 retail chains from Equinix to AWS. The CTO and CISO had no tolerance for risk. The solution had to be secure, include row-level security, and be fully encrypted. To de-risk, KRS worked with AWS Data Lab to create a data architecture, solution roadmap, and prototype for major workloads.

KRS attended a Build Lab and in only four days, their team built an Amazon Simple Storage Service (S3)-based data lake that ingests data from various source systems located in their data center using AWS Database Migration Service (DMS). The team also built an ETL pipeline that processes raw data from a landing zone into a processed zone, and Amazon Athena queries that power custom visuals and downstream business reports in Amazon QuickSight. To support business reporting requirements, Data Lab also helped KRS embed their dashboards directly into their application and implement Auth0 for secure Single Sign-On (SSO) access. AWS Cloud Development Kit (CDK) was used to build Infrastructure as Code for provisioning AWS Services with ease. Several months post-lab, KRS returned for a short workshop with AWS Data Lab focused specifically on adding the power of Natural Language Query (NLQ) to this solution using QuickSight Q. 

Building upon the best practices the KRS team learned in the lab, the data platform prototype they developed, and the ease of embedding QuickSight dashboards and Quicksight Q into their application, KRS successfully launched their new analytics product – Epiphany Data Neurocenter – to production post-lab. By incorporating NLQ as part of the solution, non-technical business users across the KRS organization can use plain language to ask questions of the data like “What was my fuel margin last week?” and "Which product is performing the best on the west-coast?”

Since deploying, the team has been able to rapidly innovate and create a half dozen new products. They’ve experienced an approximate 600% time-savings and now that insights can be derived from their data quickly, they’ve been able to use this time to refocus their product roadmap towards richer ROI opportunities.