The Big Data in Spain is starting to demonstrate their capabilities to cope solutions that previously required expensive products and designs. Specifically, for the financial environment, it is already beginning to invest in different solutions depending on the business case you want to address.
First you can adopt a solution for storing large volumes of data at a reasonable price without worrying about data history to remain as such solutions such as Hadoop, not only enable the storage but also allow processing and exploit scalable information.
It is affordable to capture information from our bases of informational and operational data via a parallelized Sqoop way to store data in HDFS such that subsequently is consumable through Hive. In this way, data can be connected filled our BI tools for traditional reporting, data visualization, data discovery and data mining.
In this type of solution comes in Data Lake concept for storing data or logs of our organization, a priori, they are not necessarily consumables but may be susceptible to analyze later.
In Memory analytic
Another case would be more interesting use of information processing in memory using scalable solutions, which provides the solution to a low response time in processing and generation of actions or events according to information processed.
The technology is currently more moving for such use cases is Spark offers various components for stream or analytical processing, among others. By processing in streaming we can equip our systems for new applications of fraud, consumption patterns, customer behavior analysis to provide it with a recommendation system that fits your needs.
In turn these technologies can be combined with products for correlation event process or business rules providing the ultimate solution processing capacity aligned with business needs.
One final use case could be the analytical processing undoubtedly generate greater value to stored or processed with Big Data technologies data and the capabilities of Data Scientist for predictive and descriptive analysis will enable large volumes data or information that is generated in real time to apply techniques or consolidated in this new type of environments algorithms. In this way you can analyze customers with a 360 view giving us a better understanding of our customers that will materialize in better service.
In conclusion the different technologies that have been mentioned briefly can cover different use cases that were previously difficult to implement at a reasonable cost and is being considered big data solutions in the architectural design of new systems or migration current hardly they meet the need of the business.
As mentioned one of the main trends of Big Data in 2015 shall be aligned with the processing in memory with Spark that is gaining ground to standardized batch processing with Hadoop.
I am Mathematical and Computer Engineer with several experiences in Java environment. Actually I am working in banking environment as Big Data Architect in Technical Architecture area. I am also enthusiastic in Analytic Models about Machine Learning and Data Mining.