Hadoop Real Time

abandonedcart

Client

Large US based corporation

The client is a large US based corporation. Their transactional system generates 4TB of data daily.

Challenges

Project scope

This data needed to be processed both in real time and in batch processes for analytics, eCommerce and marketing campaign management.

Solution

A real-time ingestion platform

Datametica built a real-time ingestion platform consisting of Flume, Kafka, Camus and Hadoop/Spark.

Solution

Real-time A data processing transit
Our Solution also included:

  1. Data Processing in transit in real time at very high volume and transaction rates
  2. Allows the additional use case of real-time fraud detection

Benefits

time-1
Real-time

Ability to serve real time as well as batch processing requirements

time
Improved analytics

Improved analytics capability as the data is available for both real-time and Hadoop analytics

operational
Improved performance
costsavings
Sales increase

Maximized revenue streams, creating several business initiatives that have measurable increases in sales and margin

Top