To improve customer retention, service providers must understand and predict customer attrition—or churn. HPE, NVIDIA, and Cloudera provide a big data analytics solution for churn analytics.
In today’s growing service provider marketplace, customers want competitive pricing, value for their money, and high-quality service. Many service providers are seeing attrition of their customer base, or churn, as customers won’t hesitate to switch providers if they are not satisfied with the services of their current providers.
According to a survey, the cost of acquiring new customers is 5X greater than the cost of retaining existing customers, with customer experience being the leading factor in loyalty and retention.
Understanding the reason behind churn is vitally important to improving the customer’s experience. To better understand churn, service providers need to collect data relevant to customer behavior, effectively store and analyze the data, and have the right modern infrastructure to achieve the indicated business outcomes in less time and in a cost-effective manner.
The answer to this need is an end-to-end big data analytics solution. In order to create an end-to-end data analytics solution that supports seamless integration of accelerators to optimize machine learning (ML), you need a performance-optimized and scalable infrastructure and the right set of ecosystem components.
End-to-end big data analytics solution for customer churn analytics
HPE, along with NVIDIA and Cloudera, provide an end-to-end big data analytics solution for customer churn analytics that is modular, flexible, secure, and performance optimized. The solution can be deployed on-premises in private cloud environments and is accelerated using HPE Elastic Platform for Analytics (EPA) along with NVIDIA-Certified Systems™ running Cloudera Data Platform-optimized software to deliver the highest performance for analytics.
The solution consists of the following components:
HPE Elastic Platform for big data analytics. This modular infrastructure foundation addresses the need for a scalable, multi-tenant platform, enabling independent scaling of compute and storage. Infrastructure building blocks are optimized for different workloads such as near-real-time streaming analytics, interactive analytics, batch analytics, and AI workloads.
NVIDIA GPU-powered accelerators (A30/A100) with a RAPIDS library for data science pipelines and RAPIDS Accelerator for Spark 3.0 software. Spark has become the de facto standard for most big data analytics workloads including AI/ML application development and deployment. The combination of Spark with NVIDIA RAPIDS accelerators provides a unified GPU-accelerated pipeline for data preparation, including extract, transform, and load (ETL), interactive analytics, and AI/ML application development and deployment with accelerated performance and scale.
Cloudera Data Platform (CDP Private Cloud Base 7.1.6): This on-premises version of the Cloudera Data Platform combines the best of Cloudera Enterprise Data Hub and Hortonworks Data Platform along with new features and enhancements across the stack. CDP empowers organizations and users to store, process, and analyze the data with the help of rich ecosystem components integrated through the single management plane. It allows users to deploy seamless services to create a compute cluster to meet business requirements with security and data governance.
The following figure is the high-level architecture for HPE Elastic Platform for Analytics (EPA) for the customer churn analytics solution.
The various stages of the end-to-end data analytics solution for implementing customer churn analytics are detailed in the following list:
- Ingesting data from diverse data sources using Kafka and NiFi in near-real-time for further processing and storing data in a Cloudera data lake
- Data transformation and processing of both structured and unstructured data using Spark in scalable core infrastructure
- Persisting the data in operational data stores using Kudu, HBase, and Impala for interactive analytics
- ML/DL application development along with ETL workloads to speed up the data preparation and ML/DL pipelines with Apache Spark, TensorFlow, XGBoost, etc.
- Data lake to persist vast amounts of diverse data on Cloudera Data Platform
HPE Elastic Platform for Analytics: Key benefits
The HPE EPA for customer churn analytics solution with the NVIDIA platform consisting of GPUs and the NVIDIA RAPIDS Accelerator integrated in the Cloudera Data Platform is designed to be a modular, flexible, and performance-optimized solution. This platform supports data ingestion, data processing, data persistence, data visualization, and Hadoop data store, along with ML development and deployment for customer churn analytics with the following benefits:
- Breakthrough economics. Significantly better density, cost, and power through workload-optimized components.
- Elasticity and flexibility. Scale storage and compute independently.
- Efficiency with performance. GPU-accelerated nodes for data processing, data analytics, and AI/ML.
- Multi-tenant platform. Unified platform for multi-user along with heterogeneous platform support, with batch analytics, interactive analytics, and near real-time analytics.
- Accelerated time-to-value. Simplified deployment of end-to-end data pipeline big data cluster using Cloudera Manager and Apache Spark integration.
NVIDIA-Certified Systems from HPE enable enterprises to confidently deploy hardware solutions that securely and optimally run their modern accelerated workloads. NVIDIA-Certified Systems are configured to deliver excellent performance for a diverse range of workloads, including Apache Spark. Customers can run most accelerated applications on these systems, including GPU-optimized software from Cloudera, and be confident that they will perform well.
Get started on the path to customer satisfaction insights
HPE and its technology partners NVIDIA and Cloudera go the extra distance to accelerate success. The customer churn analytics solution from this collaboration will deliver insights to boost customer satisfaction and retention for optimal economics.
Ready for more? Check out these additional sources for more information. And then contact us!
Solution brief: HPE Elastic Platform for Analytics for Customer Churn Analytics
White paper: HPE Reference Configuration for Elastic Platform Analytics (EPA)
Cloudera Data Platform white paper: HPE Reference Architecture for Cloudera Data Platform
Meet Compute Experts blogger Bhuvaneshwari Guddad, Solution Architect
Compute Experts
Hewlett Packard Enterprise
twitter.com/hpe_compute
linkedin.com/showcase/hpe-servers-and-systems/
hpe.com/servers
ComputeExperts
Our team of Hewlett Packard Enterprise server experts helps you to dive deep into relevant infrastructure topics.