I am attending the IBM Big Data Management launch today and will do my best to blog it as it happens. Steve Mills, on video, kicked things off.
Steve began by pointing out that the price performance boost of recent years has made a whole new class of use cases reasonable. In particular more data is being analyzed more quickly all the time and this is creating new opportunities as it becomes possible to rapidly ingest, filter, analyze and use huge volumes of data of different types. IBM he says has a whole range of technology targeted at “Big Data” and how to use it.
Dave Laverty followed on to introduce Bob Picciano the GM of IBM’s Information Management group. Dave took a minute to point out that IBM’s Almaden research facility, where the event is being held, has been the source of many research innovations from the hard disk to SQL, relational databases and moving individual atoms around! Today the focus is on innovation in Big Data.
Bob summarized the key announcements for the day:
- DB2 with BLU Acceleration – faster reporting and analytics, less disk space, included in the next release.
- Big Data Platform Enhancements – new version of BigInsights integrating enterprise-ready Hadoop, larger scale analysis for streaming data, governance, SQL access to Hadoop data.
- PureData System for Hadoop – appliances with Hadoop, archiving and built-in analytics accelerator
These new releases are targeted at customers facing new threats to their business models – demanding and connecting consumers, brands being built and destroyed in days and the growing realization that strong relationships are trumping strong products. All this fueled by the disruption caused by the explosion of Big Data, the “next natural resource.”
Leaders, he says, are leveraging Big Data to deliver immediate value at the point of impact across both operational systems and through systems of engagement. As the generations shift and more consumers are more comfortable with their data being “out there” and shared, this will only accelerate. The most effective way to use this data, he says, is to augment existing operational systems (though he still implied this has to involve presenting it to people not driving systems behavior).
IBM has identified 5 getting started scenarios:
- Enrich your information base by analyzing more data faster
- Improve customer interaction with richer view of customer
- Reduce risk and prevent fraud in security/intelligence
- Optimize infrastructure and monetize data
- Gain IT efficiency and scale with better data infrastructure
And these apply to most industries.
To get value, Bob says, you need a Big Data platform. A platform has to bring streaming,data data warehouse capabilities and Hadoop together under a unified integration and governance approach and combine all this with accelerators for discovery, application development and systems management.
The role of analytics in getting value from Big Data is pretty clear and IBM is also working on how to integrate its Business Intelligence and Analytics tools into this platform. The line of business needs analysis for its tactical and strategic decisions, decision management for its operational decisions and supporting discovery and predictive analytics capabilities. This requires support from “traditional” data warehouse infrastructure as well as raw data for data mining and analytics and streaming data. And all this data has to be ingested and managed.