Adam Gartenberg's Blog

Business Analytics and Optimization, IBM and Social Marketing

Live Blogging the Information on Demand Conference General Session


Live Blogging the IOD 2012 General Session.  (I'm posting my rough notes for now, and will look to update it later.)

Host Jason Silva:

  • Co-evolution of humans and technology.
  • Accelerated evolution.
  • Convergence of Big data, analytics, social , mobile, cloud.
  • 90% of the worrld's data created in past 2 years alone.
  • We have a voice, and we leave a digital trail.
  • We can see patterns that we cold never see before.  We can spot trends as they're happening.
  • "We're approaching a new renaissance... and all you have to do to participate is to think big."
  • "We have the tools - all we have to do is imagine what could be."
  • Digital nervous system - Make organization's feedback loop entirely digital.
  • "Big opportunities to transform businesses, industries, and the world we live in.  This, my friends, is an upgrade."

Robert LeBlanc, SVP, Middleware Software
Smarter Analytics in the New Era of Computing

Study of CEOs - 1800+
What factors are impacting organizations - Since 2004, technology has risen to #1 factor (now above people skills, market factors, globalization.)

CIO survey - What's driving innovation and market growth?  Big Data Analytics (83%), followed by Mobility - 74%, Cloud 68%.
Aren't seeing differences by industry or geographic location.  Used to see different factors in US, for example - no more.

We are embarking into new era of computing.
Three V's
  • Volume:  T-Mobile:  17 billion events a day.  Want to be able to analyze it in real time to provide better service to their clients
  • Velocity - Dublin City Council.  Uses real time data in transportation vehicles (bus, train), providing 50 updates/second.  Allows them to adapt in real time, now one of most efficient transportation systems in the world
  • Variety - Vestas (windmills).  Analyze 2.8 petabytes of data to make sure are placing windmills in the right place.

We've seen this before.
1960's - all about transaction systems (banks going online, airline reservations)
1990 - eBusiness - changed the way we thought about applications and provide transactions
2010 - Analytics - Not just producing info, but how gather it all up, provide platform to take all sources of info and process it and get insight.  The outcome you want is insight from info.  If you can't get info and can't get itt in real time, insight won't be as good.

IBM Big Data Platform
Often think of Big Data as Hadoop.  Hadoop is a part of it, but it's much more than that.  Content in ECM systems, data warehouses.  And you want stream computing because you want to act and react in real time, not after the fact.

Dr. Phil Anno, Prinicipal Scientist, ConocoPhillips
Operations in 30 countries, largest producer of oil and gas north of the arctic circle.
US Based multinational energy company headquartered in Houston, Texas.  16,000 employees worldwide.  

About the Big Data project involved in:
About 25% of earth's remaining conventional oil and gas reserves reside in the arctic, but presents many challenges due to remoteness of arctic region and short summer region.  
ConocoPhillips has invested over $500B in region.  Just one exploratory well will cost around $350M in the arctic region.

Industry will require computational system for real time tracking of ice flows over the region.  Safeguard personnel and resources.  

With IBM have developed satellite driven real time computational technology .  Near term - extend drilling season of summer by 1 month.  Drill 2 wells per summer rather than just one, and save mobilization cost on second well of around $300M per drilling season.  Longer term, with real time predictive capability, can influence design criteria for production platforms to come.  Can plan ice management operations to deploy tugboats, icebreakers to manage the ice around facilities.  Eases requirement to design production platform to withstand maximum impact.  Savings estimated to be as high as $1B in savings per production platform.

Hurdles that had to overcome:
Challenge was to track all arctic ice flows across all high arctic leases.  Like hurricane tracking, but can be several hundred ice flows moving over the path at a given time, each with own path and own cone of uncertainty.  This tracking technology, will generate around 1 terabyte of data a day.  Developed system of differential equations governing motion of ice.  Can generate accurate forecasts of where ice will go.

What pieces had to use to build the solution:
Chose to build the solution on the InfoSphere Streams platform.  Very attractive as off the shelf, flexible stream computing platform and development platform.  Developed a lot of proprietary application software, but not in the app software business.  Streams allowed their subject matter experts to focus on the tracking and analytics modules and take for granted the stream infrastructure for real time data flow.  Could also use agile software development methodology.  Due to team effort of partnering with IBM, completed it in just 5 months.  4 months of R&D and 1 month of dedicated testing.

What are you looking at next?
Next step - look to operationalize this technology.  Add some key additional data streams to constrain the forecasting solution.  New data acquisition systems are being engineered as could speak.  Looking for additional applications outside of the arctic.  For example, real time analytics on drilling, trading of electric power and natural gas.


Robert LeBlanc
IDC:  You will spend 68% of your IT budget just on operating costs.  (Up substationally from 1996).  How free up time and money to focus on challenges in front of us.
IBM Study:  Only 1 in 5 organizations allocate more than 50% of IT budget to new projects.

Change how we think of IT so we can free ourselves to focus on big challenges in organization.  That's why we came up wtih PureSystems.  Our clients are spending too much time building out infrastructure, especially when 80% of time is spent on common tasks.

Build out systems and bring our expertise and your expertise together.

Integrated all the hardware, from the compute to the storage to the networking, all in the box and pre-integrated.  On the software side, we said why don't we prieintegrate all of the middleware.  All cloud enabled (can virtualize and provision).  

And we know you use multiple tools to manage it all.  So built single integrated management console.

Free up your time so you can be innovators, rather than building out that next system.  Want to invert the %s so you can spend 68% of your time innovating.

First two:
PureFlex - Delviering Infrastructure Services right out of the box
PureApplication - Dleivering Platform Services

Next:  PureData System - Optimized exclusively for data services.

Why did we do this?  What problem did we think we could help you solve?

Inhi Cho Suh, VP, Info Mgmt Prod Mgmt and Strategy
What challenges:  Big Data, Cloud, Mobile are placing a lot of pressure on data systems not equiped to handle them.  Companies spending time trying to deploy and tune them, rather than delvering solutions.
How fit into the rest of the family:  Has common themes:  integration by design, built in expertise, simplified experience.
For example, can build patterns, and deploy database.

Addressing the data challenge with IBM PureSystems
What companies want to do is deliver values back to business with greater speed and consistency.
When talk about data load ready in matter of hours.  From cabling to integrated compute capabilities, to firmware upgrades - all done in one process.
Data workloads - transactions and analytics.  

An e-commerce system is fundamentally different from a customer analysis system, and that in turn is different than a fraud detection system.

Announcing 3 models of the PureData system - different computer sicence techniques to optimize each ssytem
PureData for Transactions - optimized exclusively for data transactions - integrated DB2 PureScale, compute memory and storage for entire lifecycle of system.  Can consolidate over 100 database workloads into 1 system.  Savings in compression, storage and speed are amazing
PureData for Analytics - optimized for analytic workloads.  Powered on Netezza technology.  20x concurrency and throughput for tactical queries.  Including industries richest set of in database analytics
PureData for operational analytics.  Optimized for operational analytic data workloads.  Integrated InfoSphere Warehouse 10.  Can run complex data analysis even while having thousands of concurrent operational queries running at same time.

Robert LeBlanc
One of questions asked in CEO study is new complexities are facing?  Responses:  We need better information and insight; the time available to capture, interpret and act on info is getting shorter and shorter.
Does analytics create a competitive advantage?
63% percentage of respondents who cited a competitive advantage
3.6x: Organizations competing on analytics are more likely to outperform their peers.

Company that embodies smarter analytics in era of big data computing.  2600 members over 86000 non-acute health care sites.  Support over 400,000 physicians.  Award winning technology that llows members to

Keith Figlioli, SVP, Premier
Represent 40% of the overall healthcare market.  Focus on performance improvement activities to drive high quality and cost effective care using technology.

What challenges face?
US Health care market going through a major transformation.  Putting in place new workflow systems, but are siloed and information poor.  Very few systems connect data across provider settings.  Or even think about population health (e.g., all diabetics in community.)  Ex., how many of us have gone to doctor's office and filled out same paperwork, only to go to imaging center and have to fill out all over again.

Providers need advanced analytics today.  focus on the heart of that problem.  

2 years ago teamed with IBM to build next generation platform, and launched it last summer.

At core of what trying to do:
1.  Connect our data
2.  Connect our knowledge base
3.  Connect our people

Build out next gen data warehouse that includes data warehouse, social business, and workflow to drive right data and process to right people at right time.
Process over 2.5 million clinical transactions in real time.  1 out of 4 patient discharges.

View across quality, safety, labor and overall population health.

Can also run large scale industry collaboratives that can lead to industry best practices.  Have become national entity to drive large scale business networking, which is directly integrated into platform, rather than separate entity.

What value do members derive and what are the big payoffs?
Value drivers:  
1.  Leveraging integrated payer and provider model.  Have integrated waste report that shows improvement opportunities.  Est. that $750B (almost 30 cents per dollar) wasted on unneeded care, complexity, fraud, etc.
2.  Collaborative - Members have amassed thousands of protocols to drive safe and efficient care - real world patterns rather than trying to do in trails.
3.  Real time data collection and predictive rules and alerting.  Predicts variance before it occurs.  1 out of every 20 patients in US healthcare will get infection while in hospital.  100,000 die each year.  One of nation's top 10 causes of death.  

Challenge of implementing core transaction systems while also building surveillance system at the same time.  Other industries can focus on surveillance because basic transaction system is now there (built out over past decades).  Healthcare still catching up there.  Providers have no choice but to do both simultaneously.  

Robert LeBlanc
How are companies already leveraging analytics to optimize outcomes?
  • TridentMarketing - 10x increase in revenue by growing retaining and satisfying customers
  • ConAgra Mills - 5% increase n production capacity
  • conEdison - $3M savings by decreasing financial risk
  • Omnicom Group - $200M increase in cash flow by transforming financial processes

Deepak Advani, VP of Business Analytics
Innovations in Smarter Analytics

All data sources, all information, all perspectives (not just pass, but what's happening right now and what's likely to happen).  Need to poptimize for all decisions and for all people in the organization.

Integration of data that's on premises and that's outside of the organization.  Analyzing data at rest as well as data in motion. Integrating data and analytics... analytics moving to where the data is and not the other way around.  A lot of the data is unstructured, it's text, it's multiple languages.  Need to understand cynicism and emoticons.  Need to understand culture differences.  No company that has invested as much in natural language processing as IBM has.  

Working with customs and border patrol to take pictures of all license plates crossing the border to analyze patterns (how often are they crossing), and bring that together identity analytics (are they who they say they are, are they a convicted drug trafficker).  If you're going to decide to pull that car over, need to decide it in real time.

Decision Optimization:  Business Rules + Optimization + Predictive Analytics.  Launched earlier this year.  Need to have intelligence embedded at the point of decision in the business process.  So a call center can make offer without knowing a thing about data analytics.  

Analytics for Everyone:  Novel visualization.  Can design what want visualization in hours, not days.  Cognos Insight: Free.  Can go to analyticszone.com and download a version of Cognos Insight for free.  Boeing - Over 30,000 users on Cognos.  They have the dashboard on 52" touch screens so that the people on the factory floor can access it.

Analytics in the cloud:  What we are seeing is that having private cloud-based analytics service is working out well.  Organizations that have centralized center of excellence for analytics tend to be far more advanced in their analytics journey.  We're packaging up private cloud to make that easy and to share best practices from our experience and those of our clients.  There is also a lot of need for a public cloud for applications and solutions.  

Demo
Running marketing for an airline.  Dashboard:  Flight delays, revenue over time, passengers over time both declining.  From overall sentiment down to integrated view down to customer level.  Know preferences so can make more relevant offers (e.g., they tend to travel over school breaks, etc.)

Want to keep track of not only transactional data, but behavioral data, attitudinal information, etc.

Robert LeBlanc
Three things to remember as this week:  
  1. Leverage analytics as a game-changer
  2. Embrace big data to maximize your business outcomes
  3. Get inspired and take action.

Think Big: Your Business Depends on it

Dave Laverty, VP Marketing Information Management

A few of the announcements spanning portfolio:
  • Accelerate value from big data:  InfoSphere Data Explorer
  • Protect your information lifecycle and reduce risk:  InfoSphere Guardium v9
  • Improve IT economics:  PureData System
  • Enable all people to make better decisions:  Analytics as a Service, Free Personal Edition of Cognos Insight
  • Deliver actionable insight:  ManyEyes (Advanced Visualization); Risk Analytics (Basel III and Solvency II)
  • Leveraging market-leading services and research:  2012 Big Data Study
  • Improve outcomes with content solutions:  Patient Care and Insights; Intelligent Investigation Manager; Defensible Disposal