Galit Shmueli, SRITNE Chaired Professor of Data Analytics at ISB Hyderabad

First, the most important step that many organizations in India must jump is data management. Getting the data stored and organized in a way that is accessible and effective for BA is crucial, yet very few companies are in that stage.

Second, those who have been able to move up this painful step are mostly involved in ‘business intelligence’, which is focused on data management and reporting. While cutting edge reporting and data visualization tools are dramatically changing the scene, these are still not the more advanced ‘predictive analytics’.

Last, while different types of data will continue to evolve (from text, to videos, to social network data, and beyond), the basics of analytics will remain the same. Building a strong basis and keeping abreast of new applications, software and techniques will likely be the secret of strong BA-based organizations.

[divider top=”1″]

Raj Mohan KK, CEO at Satvik

Traditionally, Data means the information’s stored by enterprises themselves, such as financial data stored in some Financial applications, customer data stored in CRM applications, operational data stored in ERP systems etc. We call this as Structured data. Both machine- and human-generated online transactions have been churning out large volume of data in varied formats in high velocities. We call this as UnStructured data.

We call the combination of the Structured and Un Structured data as Big Data. This brings in enormous opportunities and challenges to organizations. Traditional data management and business analytics tools, services and technologies are un able to attend this huge Volume, Velocity and Variety of data in an effective manner. This requires very radical thinking in the way organizations would like to harness the power of this sudden influx of huge volume of useful data.

This will not be an easy transition for most enterprises, but those that undertake the task and embrace Big Data as the foundation of their Data Management practice stand to gain significant competitive advantages. By designing and implementing right business solutions making use of the power of Big Data, enterprises can gain unprecedented insights into customer behavior, potential risks, volatile market conditions etc, thus facilitating them to make fact- driven business decisions faster and more efficiently.

Solutions in Consulting to Big Data Analytics to Front-end data visualization bring significant opportunities. Those vendors that aid the enterprise in its transition to Big Data practitioner, both in the form of identifying Big Data use cases that add business value and developing the technology and services to make Big Data a practical reality, will be the ones that thrive.

[divider top=”1″]

Dr. Dakshinamurthy V Kolluru, President at International School of Engineering, Hyderabad

There a lot of new trends in analytics and a lot of hype about big data. People seem to be mixing big data and analytics. It’s important for a practitioner to understand where each specialization fits.

In the role of an analyst it’s important to identify and detect non trivial patterns in data using techniques like forecasting, prediction, classification, etc. With internet, the kind of data that needs to be analysed has changed so much that there’s a necessity to have a new field whose job is to compile and collate the data before the analysts start to analyse it. The amount of data generated by Google or Facebook runs into several tera or hexa bytes. First, this data should be distributed in hardware, so analysis can be done in real time.

There are two things to be considered here: (i) To put data on hardware and then (ii) The analysis in near real-time. Classical big data engineers specialise in technologies like Hadoop, MapReduce, etc. They focus on putting together the data and structuring the data so analysis can be done in real-time.

These are two distinct skills that are needed to handle a set of problems effectively. Hadoop and MapReduce are used to engineer the data for analysis. Decision trees and neural nets can be used to solve analytics problems. Today, the focus is more on finding out the right algorithms and working it out rather than just a tool based approach.