Srirangpatna talked about some of the positive aspects of Big Data including fraud detection, pricing optimization, weather forecasting. He said we appreciate real time analysis that saves us time and cost especially when it can involve credit card fraud. For example algorithms can track your regular purchase geographic area , stores and alert you if behavior deviates from the regular patterns saving us unwanted charges and claims. Next he provided an example of how small business can use data to operate more profitably. Data driven firms are 2.5x more profitable compared to their peers. Most firms use straight line pricing with variations such as peak price, non-peak price. He said firms can refine the pricing to improve the pricing matrix to have more variations based on customer purchase patterns. This consequently improves profitability. They can also use big data to deal with non-peak demand such as hotel rental in off-season or post holiday week to generate demand and improve occupancy. Firms can use big data to improve marketing, operations and sales. He recommended attendees to try out free analytic tools such as Google AdWord, Google Analytics, Facebook Insights, LinkedIn Analytics and to scale upto commercial or paid solutions based on budget and data. Finally he said, start small -with the most important database, gradually add other databases and sources based on comfort and experience with the big data tools.
Dave Winters, CTO, FrostEdge led the second advanced session. Frost Data Capital is a venture capital firm specializing in big data analytics and health technology. Winters has consulted on Big Data projects for firms such as Facebook, LinkedIn, EA etc. Previously he has worked for firms such as Teradata and Informatica. Winters started off the audience with a teaser on a Big Data – issue with transferring huge amount of data – power and cooling. He educated the audience that security, data transfer design, encryption, replication, synchronization are some of the technical challenges Big Data users face in the enterprise warehouse. He highlighted the process to create Big Data: load, profile, parse, transform, cleanse, match and finally re-load. He talked more about structured data and unstructured data and data storage in the context of Hadoop (Big Data database). He said business users are demanding more, and want specific proprietary solutions in the context of the industry and vertical. With firms asking IT to do more with less, and speed of technology changes, IT is often found wanting to keep in step with business transformations or disruptions. He encouraged firms to be progressive with enterprise Big Data and be futuristic. Finally, he said “data” will be the new currency and consistent data quality after applying the Big Data process will break down silos within the organization.