Table of contents for IBM IMPACT 2014
- Mobile and Cloud at #IBMIMPACT 2014
- More on mobile at #IBMIMPACT 2014
- Made with IBM Keynote #IBMIMPACT 2014
- The Mobile Enterprise: Driving the Need for Better Decisions #IBMIMPACT 2014
- A cloud panel at #IBMIMPACT 2014
- Data Technical Panel at #IBMIMPACT 2014
- IBM General Managers Panel #IBMIMPACT 2014
- Smarter Process Update #IBMIMPACT 2014
- Predictive Analytics and Decision Management Video Interview #IBMIMPACT 2014
Some data-centric technical experts participated in a second panel at IMPACT. There were a number o data-centric announcements at IMPACT and they quickly recapped these announcements:
- New Power Systems
- Real-time actionable insight – capturing events and data at high speed, predicting probabilities from this data, driving this into decisions by embedding existing their Decision Management stack into streaming data.
- BlueMix data and analytics services, some focused data storage/acquisition in SQL and NoSQL formats and some analytic services – geospatial analytics, predictive etc.
- BlueInsight using the Catalyst Insight technology for rapid discovery on the cloud when combined with a data refinery.
- Catalyst Insight is also being developed for on-premised use, focusing on making it easier to develop predictive analytics – putting a data scientist in a box if you like.
Lots of interesting questions and answers. Some specific ones:
- An interesting topic is how Catalyst Insight and Watson Analytics overlap or compare. They share some components, the piece that finds the interestingness of the underlying pieces. They are seen by IBM as a pair, complementary and solving related problems.
- Context computing is an phrase IBM is going to be using more and it is especially important in real-time – picking up sensor data and deriving a usable context. Many systems only use a fraction of the available data especially when it is streaming, The new environment allows IBM’s portfolio of analytic and decisioning capabilities (rules, predictive analytics, optimization etc) and inject them into a streaming environment so that this streaming data can be acted on immediately.
- Visualization and end-user tools are challenging in this new environment. As data moves quickly the kinds of visualization that work change with many traditional visualizations being focused on a static data set. Plus, of course, as you expose these capabilities to folks with less data science expertise it becomes important to expose explanations, describe the meaning of things etc, not just displaying pictures.
- To manage all these new streaming data sources IBM is working with sensor vendors to interface their data streams and has a kit for others to develop your own integration for sensors.
- IBM has found some common patterns around Big Data and is developing these patterns to help customers to get up to speed quickly. For instance enhancing the customer view, data discovery, security intelligence, operations analysis & data warehouse augmentation.
It will be interesting to see how customers adopt streaming technology but good to see how many of the existing analytic/decision management products are being integrated with the streaming infrastructure.