Meaning Based Computing: Managing the Avalanche of Unstructured Data

Transcription

1 Meaning Based Computing: Managing the Avalanche of Unstructured Data Bill Tolson, Director of Information Governance POWER PROTECT PROMOTE

2 Moving away from structured data

3 Accelerated unstructured information growth Proliferation of Personal Devices Growth of , Collaboration & Social Networking in the Enterprise Every user in an enterprise is now an information producer and consumer resulting in: Increasing costs of information storage Difficulty accessing specific information End user expectations not being met Every two days we create as much information as we did from the dawn of civilization up until 2003, Eric Schmidt, Google

4 Content is now also, interactive POWER PROTECT PROMOTE Growing volume of traditional content Internal and external lines blurred Business use of social media Moving to interactive content

5 Increasing complexity of unstructured information - Convergence of Physical and Digital Information - Information is everywhere on paper and tapes, on edge devices, onpremises servers and now in the cloud exabytes of new data expected in 2011 (Kahn Consulting 2010) billion s per day (Kahn Consulting 2010) - Information silos Proliferation of media types and formats - Globalization of business Distributed architectures and organizations Virtualization Business velocity RTO/RPO expectations are tightening up Roughly 99% of data older than 6 months is never looked at again

6 Unstructured information is risk - More intrusive regulations 20,000+ records retention laws Higher expectations to be able to find information - quickly Tougher sanctions for failure to comply (including incarceration) Reputational risk for high-profile breaches Growing understanding by regulators and legislators of what technology can achieve - Enterprise information a target of legal discovery erecords defined - FRCP Rule 26; ESI & metadata are records Spoliation judgments are growing There were more ediscovery sanction cases in 2009 than in all years prior to 2005 combined (Duke Law report) Defensible capture and destruction practices are needed to reduce liabilities

7 Evolving ediscovery challenges POWER PROTECT PROMOTE Includes all electronic content: even social media, audio, video and SMS texts Risks and costs associated with handing off client data between systems Must locate, access and manage all enterprise data Greater number of multi-party litigation and multilingual cases Growing judicial scrutiny and larger, more punitive spoliation sanctions The average cost of a single case can easily exceed $1.75 million. Gartner On average, businesses spend $4 million per year for a $1 billion in revenue company. Cohasset

8 Information management pain points for organizations Inconsistent Retention and Disposition Poor access controls Secure Storage Ad-hoc Discovery Storage sprawl Over-preservation of information Increasing litigation risks Information leakage What is a record? management Expensive to Access Information Inadvertent deletion High Costs Weak Chain-of-Custody Manual Information Access Failure to recover from failure CAPX of Disk for Backup Unbounded Discovery costs Inefficient Business Processes Complex backup process Unprotected Laptops Catastrophic failure response Ad-hoc tape management Service continuity Unable to extract business value from data Poor customer service Information Silos

9 Evolution: need for automated records management POWER PROTECT PROMOTE ALL information must be managed, not just records Manual methods are not scalable nor consistent Evolving regulatory and statutory requirements are causing constant process changes Greater reliance on cloud solutions Ability to manage in place is needed

10 In the face of an ongoing electronic information explosion, information management becomes an imperative. Organizations need to understand how (and why) information flows throughout their organization from creation to destruction What information does your organization really have? What information should you keep? What should you dispose of, when? Are there laws that specify what information is kept and for how long? And, when you need specific information, how can you find it quickly and easily?

11 Why should any organization care about unstructured data management? Reduce the cost of storing information Reduce the cost of using information Lower the risk of not meeting ediscovery responsibilities Lower the risk of not meeting compliance requirements Raise the security of protected customer and employee data Raise the protection level of corporate intellectual property

12 Information management: easy concept, difficult to accomplish In large corporate infrastructures, information can be stored anywhere, in many forms and revisions, with little or no central control On-Premise Infrastructure Employee-Controlled Information Physical Cloud Based Information

13 An example of the problem MB new sent and received created each day for each user Users GB of new volume created each Month TB of new created each year

14 Employee managed information is the hardest to manage 1. Receive and create information 2. Information is stored in one or more storage locations Create - Receive 3. Employees use and re-use information Destroy Store Backup 4. Employees send/forward/copy information so others can also use 5. Information is managed(?) for later use 6. Information is determined to be valueless and is removed from the infrastructure - maybe Archive Manage Share Use Now consider this cycle happens with every single employee

15 An example Employee has 24GB on local hard disk and removable media 22,000+ employees world-wide 528 TB of employee controlled information, company-wide Is the data your employees hold locally, useful or a risk to your organization? Compliance requirements Discovery requirements IP security Storage requirements

16 What s the true cost of information? At its most basic, companies exist to generate and use information to produce revenue and profit What constitutes the cost of generating and using information? Employee fully loaded salaries Data infrastructure; workstations, laptops, storage, internet, support Everything an organization supplies to make the employee useful

17 The ROI of information What is the ROI of information generation? Revenue per employee? Profit per employee? What could it be if you managed your information more efficiently?

18 Problem #1: What information do I keep? Do you have a realistic information retention schedule? Who or what makes the real-time decision to keep or delete? Who or what decides how long an item should be kept and where? And how do I ensure information I don t want is actually deleted? These are the most basic questions organizations must be able to answer. If you rely on employees to make these decisions, have you trained them correctly and have you included an audit process to ensure they are following the policies correctly and consistently?

19 Where is your organization in information management capability?

20 Total cost of management of information (TCM) Reduce the cost of owning / storing information by including intelligence to help with the millions of decisions that must be made daily Instead of keep or delete everything or one size fits all Critical Information Business Continuity Compliant Data Long-term Retention Risk-Relevant Data ediscovery Transactional Data Managed for Use Look Into Enterprise Information Relevant Information , File Servers, Laptops, SharePoint ECM, Images Delete Look Across Intelligence Cost Effective, Tiered Storage

21 Problem #2: Can I find the Information when I need it? Can you find a specific piece of information quickly? Can you find all information related to a specific topic or concept? No matter where its stored? Saving information is easy, finding specific information is very difficult. Retained information you can t find and use when you need to is valueless and sometimes a liability

22 Intelligent management of enterprise information Really knowing what you have Reduce risk associated with information and enhance compliance Reduce the total cost of managing information Enable true sharing of information Maximize the business value of the information Policy Creation Look Into Policy Information TECHNOLOGY Metadata Extract Metadata Look Across Intelligence

23 Possible Information Management Strategies

24 Traditional information management and access Scattered Inefficient Unproductive Manual Legal Hold Manual Silo d Search Archive

25 ECM - Move and manage all records into one repository Employee-Controlled Edge Data On-Premise Infrastructure Cloud Storage File Servers Servers SharePoint Servers Other Systems Is this a record? No Yes ECM Systems

26 Store and access in place utilize existing system indexes Employee-Controlled Edge Data On-Premise Infrastructure Cloud Storage File Servers Servers SharePointOther Systems Servers Index Index Index Index Index No indexes available Search capabilities of each application are inconsistent

27 Store and manage in place common index Employee-Controlled Edge Data On-Premise Infrastructure Cloud Storage File Servers Servers SharePoint Servers Other Systems Management Layer (for search, retention management and ediscovery) Super Indexer Index

28 POWER PROTECT PROMOTE

29 POWER PROTECT PROMOTE

30 Meaning-Based Computing POWER PROTECT PROMOTE Files IM Audio Social Video SharePoint Gain control, manage cost Understand Meaning Govern Information Rich user search Automatic records classification Advanced discovery & investigations Supervision Surveillance

31 Comprehensive meaning based computing for governance

32 Move away from manual processes with understanding & automation Notes Database Document Management News Feeds Information Theory and Bayesian Inference Integration Through Understanding Process Automation Manual Processes Aggregation Aggregate content, tag & categorize Automatic categorization Hypertext links to similar content Hyperlinking Personalization from forms/questionnaires, Profiling geodemographic profiling Personalization Searching for information Collaboration Internet Files XML Delivery Answering customer Retrieval inquiries via a helpdesk Reformatting for multichannel delivery, e.g. Routing PDF to XML Alerting ing information to relevant recipients Audio/ Media

33 Thank You Chris Fellers: Bill Tolson: POWER PROTECT PROMOTE