Made with IBM Keynote #IBMIMPACT 2014

Marie Wieck came back to discuss how to create the kind of composable business  IBM discussed on day 1. It’s critical, she says, to keep customers at the center and drive a new process from this. It’s key to use data to make better decisions, and it’s important to have the kind of scalable infrastructure you need to make this scale and deploy effectively.

Once again the focus is on IBM customers and HH Gregg came on stage to talk about their journey. As a regional retailer they need to need to differentiate themselves. In particular they have focused on providing a “friction free” shopping experience on every channel. They have historically invested in training for their staff and have tried to bring this kind of personalized service to their mobile strategy.  They determined that it was not enough to “mobilize” their existing interface, it had to reflect the different form factor and interaction patterns. Mobile had to lead not just follow. The customer-centric approach at the heart of a mobile strategy is becoming a driver for HH Gregg’s business transformation – the mobile strategy is driving their overall strategy.

Marie came back to bring the focus to integration. She emphasized that integration is not just a technical issue – not just about connecting to systems and data sources but about integrating your business into the new economy. A typical mobile app, for instance, will embed 6 different APIs.

A recent Institute for Business Value study looked at high performing organizations. They found that high performing organizations were 35% more likely to say that process optimization was critical. High performing organizations are 43% more likely to use social networks to manage knowledge. They are also 50% more likely to establish and emphasize a culture of innovation and change. Finally high performers were 86% more likely to see technology as a key enabler. IBM, of course, sees itself as a great provider of all these elements.

Another client, next, Coriell Life Sciences, came up next. Coriell Life Sciences uses DNA analysis to try and tailor medical treatment. For instance, elderly patients take perhaps 20-22 drugs and the likelihood is that many of these are not helping or are mis-dosed etc. Coriell uses the dramatically reduced cost of sequencing a genome – down to $1,000 – to sequence an individual to drive personalized medicine. Coriell uses the IBM Softlayer deployments, of IBM BPM for instance, to connect to its partners (doctors, labs etc) across the medical industry. Specific mapping of genes to known drug issues generates a decision support tool for the patient’s doctor and securely routes it back to them. Moving forward they are turning to IBM Watson to bring together a patient’s behavior, drug interactions, genetic sequence-based advices and more to show doctor’s the impact of different regimens.

Bob Picciano came up next talking about data and analytics. Social media, mobile devices, the internet of things are all going to push the amount of data every higher. Time compression is a reality for most companies, pushing them to real-time responses. More data, arriving more quickly, combined with a need for faster responses means that its essential to use big data in real-time analytics – decision management in other words.

Caroline McGregor joined him on stage talking about the use of fast moving, high volume data in medicine. I have blogged about her work before and it’s a great example of not just doing analytics to predict problems but of making these predictions against streaming data and then automatically consuming those predictions in automated decisions. Her work is now being extended to other potential problems in neo-natal care and even to the real-time monitoring of astronauts.

Bob came back to talk about IBM’s support for decision management – sensing what is happening, building analytics against this data and then mixing these analytics with business rules to decide what best to do before applying these decisions continuously, in real-time and at scale. BlueMix can provide all these capabilities on the cloud. The services available include MapReduce, BLU Acceleration (see this review), Cloudant and more. Geospatial services, a time series database and predictive scoring services are coming next. Plus data governance and management are provided to make sure companies can keep track of all this.

Moving into the future, IBM is delivering something codenamed BlueInsight containing a data refinery and catalyst insight. The data refinery supports data extraction, matching and improvement services and lets companies get access to these services. Catalyst Insight uses machine learning to find out what seems important in the data you refined to show business users, and developers, to see what’s  hidden in their data (see this post for a little big on the underlying SPSS technology Analytic Catalyst).

Steve Mills came up to focus on how much broader the discussion at IMPACT has become, moving well beyond the original focus on WebSphere as a platform. Steve changed gears on us, focusing back on to the infrastructure (especially mainframe infrastructure) that underpins all these ideas (one of his favorite topics). Infrastructure continues to evolve, moving toward increasingly dynamic optimization and federation built on increasingly hybrid cloud/on-premise infrastructure stacks. IBM continues to invest in improving the value offered by its mainframe platforms as more MIPS are added every year and run continuously, delivering massive amounts of work around the world. IBM is providing more analytic capabilities that can connect directly to mainframes to eliminate the huge amount of data movement that still consumes mainframe resources. Storage, Power Systems and other IBM hardware,  appliances and platforms are continuously refined to adapt to new workloads, challenges like mobile, new operating systems usage, more parallelism and more. It’s essential, he says, to consider the total cost of OWNERSHIP not just the total cost of acquisition. Doug Balog followed Steve to continue this focus on infrastructure, talking about the next generation of Power Systems and how they support the open stack that is increasingly the focus of developers as well as cloud.

Walgreens came up to talk about their infrastructure and their move to PureApplication but I was more interested in the scale of their decision management systems – rules based drug interaction decisions for instance, optimizing the scheduling and workload of pharmacists, handling 11M customer interactions every day.

Cross-posted from JTonEDM.