A Federated Approach to Big Data IBM Watson
A Federated Approach to Big Data -- IBM Watson Explorer Presented by: Ken Holmes
“Data is the new Oil. Data is just like New crude. Oil” It’s valuable, “Data is the but if unrefined it cannot really be used. ” – Clive Humby, Dunn. Humby “We have an economy based on a resource that is not only renewable, but self-generating. Running out is not a problem, drowning in it is. ” – John Naisbitt Exploration can be a critical step! 22
The demand for Big Data solutions is real The healthcare industry loses $250 - $300 billion on healthcare fraud, per year. In the US alone this is a $650 million per day problem. 1 One rogue trader at a leading global financial services firm created $2 billion worth of losses, almost bankrupting the company. $93 billion in total sales is missed each year because retailers don’t have the right products in stock to meet customer demand. 5 billion global subscribers in the telco industry are demanding unique and personalized offerings that match their individual lifestyles. 2 Source: 1. Harvard, Harvard Business Review, April 2010. 2, IBM Institute for Business Value, The Global CFO Study, 2010. 3 3
Polling Question 1. How much of available information does a typical organization utilize? A. 6% B. 12% C. 24% D. 48% E. 60% 4 © 2013 IBM Corporation
Only Of Available Data is Used Forrester Research: Can You Give The Business The Data That It Needs? by Michele Goetz, November 13, 2013 32% of structured 8% of unstructured
Big Data Exploration bridges the gap between structured and unstructured data, cloud, on-premise and external Big Data Exploration Stream, Process and analyze Big Data Federate, discover and navigate Big Data sources Virtual Integration Lacks structure Web RSS Feed ______ Social Media World’s Total Data: Content Mgt Systems Unstructured docs 80% Unstructured ERP CRM SCM SOA, ESB, Web Service Enterprise Systems & Content Stores 20% Structured Each system has its own but different structure
Polling Question 2. If “data is the new oil, ” what’s the first step in exploiting it? A. Start drilling immediately B. Build a refinery C. Open a chain of gas stations D. Explore to find the richest deposits 7 © 2013 IBM Corporation
Understanding Big Data is critical to success Explore § Discover and navigate all Big Data repositories – internal and external sources Analyze § Analyze and compare trillions of data records from structured and unstructured sources Understand § Correlate & combine all data sources to unearth unique relationships Getting Started is Crucial 8
Phase 1: Leverage Information In-Place Phase 2: Automate and Expand w/ Big Data Analytics Business End User Data Scientist / Analyst Discover & Navigate Advanced Analytics Watson Explorer Identify Analysis. Ready Data Big Data Analytics & ed d ture ruc c St nstru U In t Ex ern te al rn & al Hadoop System Critical Information The Virtuous Circle of Information Analysis Serve up Analytical Insights in Context 9 Stream Computing Data Warehouse
Big Data Exploration helps to leverage ALL the data TRADITIONAL APPROACH Analyzed information BIG DATA APPROACH Analyzed information All available information Analyze small subsets of data Analyze all data © 2013 IBM Corporation
Big Data is everywhere, but not integrated in a single view Sensor and machine data E-mail and collaboration Applications and ECM Systems of record Internal Content Contact Center R&D Hadoop Analytics ? Marketing Data Scientist Web Social media External Content Security “I can’t find the right answers fast enough to support my customers. ” “I can’t unlock the value “Innovation is falling in my data to drive short as I am unable to economic value to my see the full research business. ” picture. ” 11 Mobile data “I am monitoring all angles – yet I can’t connect the dots. ” “I can’t get my arms around all of the data in our enterprise. ” “I don’trust the data we’re using for important decisions”
Unlock the value of information when users need it most Watson Explorer Data access & integration • Index structured & unstructured Unified access and data—in place fusion across all sources • Support existing security • Federate to external sources Improve customer service & reduce call times 12 Analyze customer data to Increase productivity & unlock true customer value leverage past work increasing speed to market Discovery & exploration • Unified view of all information • Information-centric applications • All at big data scale Create unified view of ALL information for real-time monitoring Identify areas of information risk & ensure data compliance
Business and IT Drivers for Big Data Exploration 1. You must invest a little time to understand the critical business imperatives Understand the organization’s critical business imperatives and potential use cases • Revenue attainment • Cost control • Customer loyalty • Productivity • Compliance • Competitive advantage • Employee engagement Sample Business KPIs 1. 2. 3. 4. 5. 6. 7. 8. 9. Average order value Profit margin Net Promoter Score Lifetime customer value Gross margin Customer acquisition rate/cost Average sales price Cross sell & up sell Category margin 2. Understand Benefits of Big Data Exploration – Do you have a way to explore important data sources? Do you have a way to deliver big data & analytics to the employees who need it? Explore all data to determine what is relevant to big data initiatives § Enterprise systems § External data § “New” data (sensor, etc. ). Reduce cost of big data integration § Create integrated views for new insights § Reduce time-to-value § Better use/re-use of info Deploy apps to deliver data & analytics § Improve customer service § Increase productivity § Improve employee performance © 2013 IBM Corporation
Benefits of big data exploration are felt throughout the organization—some examples Sales § Less time looking for info; more time in front of customers § 360 view of customer § Faster response to new opportunities § Better up-sell/cross-sell Manufacturing § Supply chain visibility § Access to R&D data § Improved collaboration Support § Single point of access for all info § Reduced average handle time § Improved customer satisfaction § Improved morale and retention § Increased up-sell and referral 14 R&D § Reduced time looking for info § Better re-use of prior research § Increased collaboration/expert identification § Increased innovation & return on R&D investment HR § Higher morale & engagement § Lower churn/turnover § Knowledge transfer from senior staff § Reduced training/on-boarding time Executive § Decisions made with better information § Reduced risk § Multiple ways to critical business issues Big Data Exploration use case offers multiple value propositions depending on • Where the client is in big data journey • What business issues are top-of-mind © 2013 IBM Corporation
EXAMPLES © 2013 IBM Corporation
Watson Explorer for Enterprise Reporting – 360 The Challenge The Solution Degree View For many years, this automotive industry leader led the way in understanding their customer loyalty and profitability using traditional reporting tools. Their challenge has been in analyzing unstructured data in context with their traditional BI tools and empowering ALL their business users to “see the whole picture. ” Watson Explorer Capabilities: • Connectivity framework • Powerful Text Analytics • Security model • Application Builder Their requirements: • • • Discover insights into unstructured text Scalable Platform Secure Rich connectivity In 2 weeks, Watson Explorer connected to Netezza, OBIEE, Fileshares, Share. Point to: • Aggregate Contextual Data Increase information access Increased Visibility into high value data This project is expected to net a cost savings of $7 m in their call center operations alone within 2 years. “We never thought we would be able to see ALL of our data in one place, and it took two days”
Watson Explorer Customer Example – Leading Analyst Firm Adapted relevance lead to greater user satisfaction, click through and up -sell Profile-based suggestions Reports, analysts, and other types of contents are searchable Rich navigation through faceting, clustering & related content Use Watson Explorer as an “application development platform” Integrate different content types
Watson Explorer Customer Example - Airbus Custom web applications Improved customer satisfaction and lowered costs § Problem – Provide uniform information access platform to develop multiple customer centric applications for Support, Service and Self-service – Information locked into multiple data sources with different security schemas § Solution – Provided connectivity to complex repositories such as Aqualogic, SAP R 3 and KM, Siebel – Extract and index all metadata – Supported existing security policy – Run-in parallel parsing agents § Results – Indexed 2 PB of data – Deployed in 1 month – Multiple front-end applications leveraging common back-end infrastructure – Single point access to all repositories 18 Watson Explorer • Supplier Information • Service manuals • Customer profiles • Lessons learned • Customer call details • Sales pipeline
From challenges to opportunities Leading Medical Equipment Supplier A leading medical device manufacturer delivers detailed knowledge about customers and products to their contact center agents to enable better engagement and asking “one more question” to increase cross-selling. Business outcomes • Increased revenue and decreased cost in the call center • Increased customer satisfaction & employee engagement • Created opportunities from each customer interaction - “one more question, ” targeted to individual client situation Product Starting Point: Watson Explorer 19
Large Investment Bank Need • Reduce risk and improve compliance Quote from the bank: “We knew there had to be a better way than monitoring all those different applications oneby-one. ” • Too many silos to monitor across multiple LOBs—needed visibility to all from a single point • Improve knowledge sharing and research • Deliver 360º view of customers, products and assets Solution/Status • Suite of IBM products provided solution that no other single vendor could match • Watson Explorer deployed in: • Po. C in Consumer Client Banking • Risk and compliance solution • Asset management group Po. C • Call center knowledge management • Big data initiative
Big Data Developers @ § FREE § 1000+ Members – All types of practitioners – All skill levels § § § San Jose & Foster City Evenings, Full Day, & Afternoons Hands-on Labs Live Streaming Past Meetups: – Hadoop – Text Analytics – Real-time Analytics – SQL for Hadoop – HBase – Social Media Analytics – Machine Data Analytics – Security and Privacy meetup. com/Big. Data. Developers NEXT MEETUP: Real-time Analytics Developer Day on Thursday, April 17 Coming Soon: Big Data & R, Watson, Cloud, Mongo. DB & more!
Thank you
- Slides: 22