Case Study: Retail Client
Our Client is the leading independent provider of infrastructure software creating event-enabled enterprises to use on-premise or as part of a cloud.
- Client USA’s # 1 leading Retailer was looking for a solution to provide better recommendations for visitors to their website.
- Their online competitors like Amazon and Netflix, were deriving bigger sales volumes by presenting the customers with right product/content. Client was looking to revamp their recommendation engine as the current one lacked the capabilities to provide personalized recommendation to visiting buyers based on their purchase history and products searches.
- World’s # 1 Leading Retailer’s existing systems had all their customers search and purchase data system being logged into an Oracle databases as blobs
- High Data volumes and heavy volumes and unstructured data were their biggest challenges
- Due to lack of proper data warehouse and proper customer behavior analytics there was a huge business gap and their online sales were falling short of their online competitors.
- Their Analytics team needed a better structured data warehouse system to provide insights into customer behavioral and buying pattern. There was a business need to restructure the existing database system which would provide better insights into customer behavioral and buying pattern for analysis and to showcase relevant similar products to customers to drive up their online sales enabling a better customer experience.
- Openmind’s BI Architect team and experts helped the retailer build the solution using Flume, Hbase, Hadoop and Tableau.
- We started by creating a Datamart that would source the Web events directly from the Web logs using Flume to source the relevant Data into Hbase.
- The data was then extracted into Hadoop ecosystem and then cleansed, processed, split into analytical structures to support Tableau dashboards. We estimated the size of the yearly data to be 70 TB and our team successfully provided Hadoop/Hive as the preferred solution to handle big data volumes.
- Openmind team built the successfully delivered the proposed solution in 6 months estimated timeframe.
- This solution became the core part of the critical Web analytic infrastructure. This helped our client to experiment online on their site in real time as well as the ability to measure the results instantly as opposed to earlier wait period which was over for a month. The retailer could now offer better personalized content to their online buyers enhancing their online presence.
- They were now in a strong position to offer more targeted products based on customer’s buying and search history as opposed to earlier generic recommendations. This in turn helped them increase their Online Sales and Presence.
Technology Stack Used:
- Backend: Hadoop, Hbase, Hive
- Tools/languages: Flume, Python, Hive, Linux, Scala
- Reporting: Tableau