Adam Gartenberg's Blog

Business Analytics and Optimization, IBM and Social Marketing

IOD Day 3 General Session


Rounding out my IOD 2013 General Session coverage, here are my notes from Wednesday's General Session.

The General Session began with a panel discussion of four IBMers whose work with clients across various aspects of the Big Data and Analytics spectrum, and wrapped up with Mark Jeffries interviewing Serena Williams about a variety of topics.

Panel:

  • Gigi Yuen-Reed - Healthcare Consulting Services
  • Vijay Dheap - Security Intelligence with Big Data
  • Chris O'Connor - Cloud and Smarter Infrastructure
  • Ian Story - Enterprise Content Management

Gigi Yuen-Reed - Healthcare Consulting Services
  • The changes in US healthcare laws will have a lot of implications.  From a data perspective, we're seeing a shift in how you view the data - from company point of view to individual point of view.  There's also a change in how risk management works [in the US] - no more pre-existing conditions or lifetime maximums.  These changes are removing many of the safeguards that healthcare insurance companies had in the past.  Also dealing with situations of no data - previously uninsured.  Looking at how turn data into your best friend.
  • IBM's Healthcare Consulting Services include teams of research scientists, data scientists, and consultants.  They have built out prototypes, and are now deploying solutions to clients.  
  • They are bringing together client data with public data - 3rd party data, surveys, etc.  Combine them with predictive models for a view into cost and enrollment.  Need to distinguish signal from noise.  Techniques from banking, and other sources.  
  • They're helping clients get early visibility, so they know what products are working or not.  They can also help them be proactive - reach out to high risk individuals for early intervention.

Vijay Dheap - Security Intelligence with Big Data
  • The cost of cyber attacks is growing at a 10% clip a year (Gartner), up to 102 attacks per weeks per company (The Ponemon Institute).  
  • Data professionals try to build bigger walls, but business needs flexibility.  Instead, they need to improve their security IQ.  Do it via observation - who are the people on your network.  IBM QRadar can monitor activity and find anomalies.  If there are any bad actors or bad IP addresses, can block them.    
  • Heuristics - e.g., was the domain just registered today?
  • Bringing together QRadar with Big Insights to find the positives while filtering out false positives.
  • http://www.securityintelligence.com - blog focused on security intelligence

Chris O'Connor - Cloud and Smarter Infrastructure
  • "Bit buckets on the floor" (e.g., today many people are collecting data the way you might put out buckets to catch water from a leaking ceiling) - we've gotten good at instrumenting and getting data, now we need to do something with it.
  • e.g., telecom providers have a lot of different pieces of technology coming together to make the network work.  They are collecting metrics on each of those of pieces, creating terabytes of log files and unstructured data.  Today that data sits in silos and often at best someone looks in individual buckets.  Need to be looking at relationships across buckets.
  • The goal is to give them tools, mathematics, capabilities to understand how they put it together.
  • Consolidated Communications - 80,000 individual streams coming across the network.  Predictive Insights product gives them visibility across the buckets and establishes normal mathematical relationships across those 80,000 streams of data so they can reduce the number of buckets that they need to look at, and to bring in a predictive posture in so that they can ensure the service stays online.  (See this video for more on Consolidated Communications)

Ian Story - Enterprise Content Management
  • Today, when disaster happens, you're much more likely to hear about it from a social network than the evening news.
  • Security First Insurance.  Using content analytics to do sentiment analysis against social data during disasters.  Identifying urgent cases and creating cases in case management system and then sending out agents with mobile devices, or allowing clients to take own documentation with their mobile devices.  Can assign it to the right case manager, triage to see who needs assistance first.  They can push out documents through IBM QuickFile, and clients can monitor using tool built with IBM Worklight.

After the panel, Mark Jeffries came out to interview Serena Williams, who was quite engaging.  It was really interesting to hear her perspective on her approach to the support, being in the spotlight, and how she makes use of technology and data to improve her game.

Some highlights I captured during the interview:
  • Mindframe is do best I can, stay positive.  Know if do the right things, will have a positive result.
  • Practice is the hardest part of the job.  Don't see the results in practice...
  • What's it like to always be under scrutiny?  When lose match, feel like it's usually she could have done better, as opposed to something that her opponent did.
  • How do you analyze competitors?  When I first started, didn't have as much data or technology to analyze opponents.  Would scout opponents, see how they play, but that's it.  Now have data - know percentages of where they like to serve or where they like to hit forehand.  Data can move tennis players and other athletes to a whole new level.  Coach is a big fan of data.  Loves to know numbers, exactly what happened, how she did - where serving, where moving, where moving faster to.
  • French Open hands IBM packs after every match.  When first got it, was just CD filled with numbers.  At Wimbledon, give the match (video) as well as data.
  • IBM now has the keys to the match (but not until players are on the court).  Can go afterwards, though, to see it.
  • Advice from father:  "You can't rewind time."  Have to move on to the next point and do better on the next point.  You have another chance.