Our client business is providing their customers with network hardware in retail, commercial, and as a service. In order to maintain the quality of these products and services, prompt and accurate reporting on the data from these devices is required at all times.
Approx. 80-90 GB JSON data is generated per day from Customer's routers. The customer's previous Router Analytics IoT application was loaded to Hadoop and relied upon Tableau querying the Hadoop / Hive data. Tableau reports were taking over 20 minutes to return results which were unacceptable to the Customer business teams.
Customer selected KPI for their expertise in Hadoop, Snowflake, and Tableau, whilst providing a blended shore team, is part of a multi-year engagement with the Customer to evaluate, develop, and support their analytical efforts across multiple regions. KPI's expertise, experience, and offshore project centers greatly minimize cost and risk for the Customer.
What KPI Delivered
In order to address the report performance issues, KPI proposed Snowflake on AWS. Snowflake's capacity and usage pricing for its Massively Parallel Performance (MPP) complemented the Hadoop Data Lake without the need for expensive alternative solutions.
This approach allows the KPI team to crunch real-time data, prepare, model, and visualize in Tableau in a short period of time with a set of orchestrated events.
By leveraging Snowflake with Hadoop, Tableau report performance has increased from 20 minutes to less than 7 seconds allowing users to do the true speed of thought data exploration and analysis.
Ad-hoc / Real-time analysis is a key feature for Big Data Visualization. By complementing the Hadoop Data Lake with Snowflake the re-architected Router Analytics provides near real-time analysis in Tableau. This solution allows Customer business users to visualize millions of records and supports growth and customer satisfaction goals for their business teams.