Phasic Systems Inc (PSI) Agile Data Governance, Design, Aggregation
Dec 31, 2015
Phasic Systems Inc (PSI)
Agile Data Governance, Design, Aggregation
2
Phasic Systems Inc• Product-based company established to tackle the toughest data
management problems.• Geoffrey Malafsky Ph.D, CEO
▫Research scientist▫Supported many agencies in their quest to access the right
information at the right time• Maribel Soto, Ph.D, COO
▫Research scientist▫As Director of SwampWorks led the transition of breakthrough
technologies for the Navy• Tim Traverso, VP Federal
▫Technical Director, Navy Deputy CIO
3
Problem Statement
•Analytics depends on consistency, commonality•Too much data dispersed over many systems•Data cannot be readily aggregated due to
conflicting meanings and lack of documentation •Cannot achieve enterprise analytics without using
new approaches and tools•Cannot transition to Cloud without new approaches
and tools for data management
4
Options
•Status quo: non-Agile data design, management and integration - expensive, time-consuming because requires brute-force
•Current tools▫Data modeling: ERWIN, Oracle, XML, Semantics▫Integration: Informatica, IBM DataStage▫Analytics/Reporting: Cognos, Microstrategy
•DataStar: efficient, cost effective because uses Agile business-centric modeling, integration with existing systems and simplified, evolutionary transition to new technologies (e.g. NoSQL)
5
• Authoritative data• Leverage existing infrastructure• Timely, Agile, flexible• Meaningful analytics
DataStar™Suite
Users Data Modeling
IntegrationAnalytics/Reporting
Solution
6
•Born out of experience – Navy, NGA, DOT, USMC •Organizations are unable to capitalize on data due to multiple
versions with conflicting meanings ▫Global Force Mgmt, DIMHRS
•Rigid methods of data management are compounding the problem with isolated business areas and processes
•Needed a solution that was ▫Rapid and flexible – used Agile technology▫Understandable to all – made it easy to learn ▫Able to simply handle all business variations – used standards-
based simplified semantics
DataStar: Agile Data Governance, Design, Aggregation
7
Capabilities
• DataStar Discovery: Agile data governance and design▫Re-add business context to data▫Flexible, common data definitions/ semantics, models
• DataStar Discovery Plus: Agile warehousing and aggregation▫Simplified, common semantics in flexible data models▫Source to target mapping for existing ETL tools
• DataStar Adapter: Managed high-performance data exchange service for XML standards▫NIEM, HL7, XBRL, …
8
9
Agile Actionable Governance/Architecture
10
Discovery Supports Governance Framework*Framework Component Discovery Support
Mission and Value
Goals, Governance, Metrics and Success Measures, and Funding Strategies
Data Rules and Definitions
Decision Rights
Accountabilities
Controls
Data Stakeholders
Data Governance Office
Data Stewards
Proactive, Reactive, and Ongoing Data Governance Processes
* Data Governance Institute Framework
11
The Semantic Moat•Semantics means data’s actual meaning
▫This is NOT the same as database metadata (table and element names)
•When is a “customer” not a “customer” ?▫When Sales, Accounting, Product Development, CEO, (SAP,
IBM, Oracle, ….) use it in their systems•Why can’t these be consolidated into one common
definition ?▫Strong disincentives to give up control of definitions, system
control, peer prestige▫Complexity of evolving integrated business-technology-
organization perspectives
12
Unified Business Model: Add Business Context to Data Definitions KORS
™• Simple list-based• Use all related
information• Predefined
template• Team
collaborative centric
• Automated analyses• Semantic
similarity• Data tracing• System
connections
13
Case Study #1 – Navy Data Strategy
• Problem: Large organization, dispersed mgmt, data is redundant, conflicting, difficult to govern and consolidate
• Previous approaches: 10 years of organizational policies and structures (e.g. Functional Data Mgrs), and warehouse development could not solve problem
• Solution – Used DataStar Facilitated Agile Governance and modeling• Results – 90 days, achieved new Data Strategy with concrete
process, definitions, metrics and widespread consensus• Customer Reaction – “We accomplished more in three months with
DataStar than we did in the past three years.“ DONCIO
14
Case Study #2 – Genentech
• Problem: Data could not be integrated from SAP system, in-house marketing, and external regulatory data
• Solution – Used Datastar Agile modeling and integration
• Results – 1 week, consolidated data with common data semantics without impacting original systems
15
Case Study #3 – Navy HR Data• Problem: 5 major warehouses for 1 million HR records. Conflicting
data, unable to consolidate, undocumented business context, brute force approaches don’t scale
• Solution – Used DataStar Agile warehousing & integration• Results – Agilely implemented common data model built on
simplified, stds-based definitions accommodating all business variations; built subject matter-based integration transforms; automatically integrated, de-duplicated, cleansed data into Authoritative data stores
• Customer Reaction – Data model accepted as standard with standing Governance renamed for it (Position-Resume Governance Board). Only consolidated data approved as trustworthy – VADM K. Moran (Ret)
16
Further Information
• PSI Sales▫ [email protected]▫888-735-1774
• Web site: http://www.phasicsystemsinc.com/prodservinfo.html▫Videos▫Technical papers▫Conference briefings