“THE DATA WAREHOUSE IS NOT DEAD!”
A PRACTICAL GUIDE TO
MODERN ENTERPRISE INFORMATION ARCHITECTURE
2
Specialist, Commercial Division
Austin, TX
ksharma@sensecorp.com
Kunal Sharma
About the Presenter
 15+ years leading complex data
transformation projects for Fortune
500 and mid-size companies
 Clean Data Practice Leader
• Mainframes
• Data Entry
• Basic Reporting
• Primitive Databases
1970s
• Personal Computers
• Business Applications
• Relational Databases
• Business Data
Warehouse
1980s
• Internet
• Centralized Data Storage
• Kimball and Inmon Data
Modeling Theory
• EDW Architecture Model
1990s
• Big Data
• Data Lakes & Hadoop
• Cloud Computing
• AI / ML
• IoT / Telematics
• Data Governance
2010s
• Broadband = More Data
• Business Intelligence
• Data Mining and
Predictive Modeling
• SaaS
• MDM
2000s
A BRIEF HISTORY OF THE
ENTERPRISE DATA WAREHOUSE
Source: Kellogg School of Management at Northwest University
THE VALUE OF CLEAN DATA
DIRTY DATA CAN LEAD TO COSTLY DECISIONS
The Impact of Clean Data
While we know that dirty water can
impact the health of people,
We don’t as easily accept or recognize that
dirty data can impact the health of companies..
The Problem of Bad Data
Building a Clean Data Practice
Establishing a Clean Data
Practice is dependent
upon a strong foundational
Data Platform
THE RIGHT REASONS
CONSIDERATIONS FOR YOUR DATA PLATFORM
Use Case Considerations
Compliance Reporting
Governed data produces certified results that ensure no miscues in both internal and external reporting
Impact Analysis
Change management can easily trace and identify any impacts to data consumers
Digital Transformation
Architecture should leverage a hub and spoke model to enable domain based micro service builds
System Replacement
Converting to a new system should leverage clean data as part of any data import activities
Growth By Acquisition
Requires a data strategy that supports a consolidated view of data across multiple data sources
MAKING THE DISTINCTION
SINGLE SOURCE OF TRUTH VS BEST VERSION OF TRUTH
Making the Distinction
Single Source of Truth Best Version of Truth
Data storage principle to always source
information from a single source
Multiple sources of similar data across
transactional systems
Enables transparency, traceability, and
clear ownership of the data
Impacts timeliness and completeness of
enterprise data
Data usage principle for a single agreed
upon view of data
Requires a governed Master Data
Management stewardship
Results in certified “trusted” data for all
data consumption needs
Utilize business rules to eliminate data
redundancy and define metrics
ENTERPRISE DATA ARCHITECTURE
BUILDING THE RIGHT DATA LAYERS
Enterprise Data Architecture
DATA GOVERNANCE
Data Lake
Operational Data Store (ODS)
Data Mart
OLAP Cubes
Defining Characteristics
• Daily data latency at minimum
• Structured by analytical consumer functions
• Semantic Layer with accompanying aggregation(s)
• Data cubes enable consumers to quickly slice, dice,
and summarize data in a presentation tool
Typical Data Consumers
• Production Support
• Presentation Tools
• Reporting Analysts
• Executives / Upper Management
MODERN INFRASTRUCTURE
THE CLOUD LAKE HOUSE
Cloud Lake House
Streaming
Mobile
Log Files
IoT
Social
On-Premises
Databases Files
Data
Warehouse
SaaS
Applications ERP
DATA SOURCES
DATA GOVERNANCE
Data Catalog | Master & Reference Data Management | Policies & Procedures
DATA SECURITY
User Provisioning | Protected Information | Network Access
CLOUD DATA LAKE
Raw
Zone
Structured
Zone
Curated
Zone
ANALYTICS SANDBOX
Data Scientists
CLOUD DATA WAREHOUSE
Data
Marts
ODS OLAP
Cubes
CONSUMERS
Data Analysts
Presentation Tools
Business Users
APIs & Extracts
CLOUD STORAGE
STREAM PROCESSING
BATCH PROCESSING
Utilize the opportunity to hit the reset button
Planning For Modernization
Data Governance is critical to your success
Avoid the pitfalls of a “lift and shift then fix” migration
Start small with a focus to maximize data enrichment
Take advantage of the ecosystem to avoid vendor lock
Thanks For Joining Us
We hope you enjoyed the presentation.
If you’d like to learn more about
The Clean Data Initiative,
we encourage you to download the full eBook.
DOWNLOAD EBOOK
www.sensecorp.com | marketing@sensecorp.com
Q&A

The Data Warehouse is NOT Dead

  • 1.
    “THE DATA WAREHOUSEIS NOT DEAD!” A PRACTICAL GUIDE TO MODERN ENTERPRISE INFORMATION ARCHITECTURE
  • 2.
    2 Specialist, Commercial Division Austin,TX ksharma@sensecorp.com Kunal Sharma About the Presenter  15+ years leading complex data transformation projects for Fortune 500 and mid-size companies  Clean Data Practice Leader
  • 3.
    • Mainframes • DataEntry • Basic Reporting • Primitive Databases 1970s • Personal Computers • Business Applications • Relational Databases • Business Data Warehouse 1980s • Internet • Centralized Data Storage • Kimball and Inmon Data Modeling Theory • EDW Architecture Model 1990s • Big Data • Data Lakes & Hadoop • Cloud Computing • AI / ML • IoT / Telematics • Data Governance 2010s • Broadband = More Data • Business Intelligence • Data Mining and Predictive Modeling • SaaS • MDM 2000s A BRIEF HISTORY OF THE ENTERPRISE DATA WAREHOUSE
  • 4.
    Source: Kellogg Schoolof Management at Northwest University
  • 5.
    THE VALUE OFCLEAN DATA DIRTY DATA CAN LEAD TO COSTLY DECISIONS
  • 6.
    The Impact ofClean Data While we know that dirty water can impact the health of people, We don’t as easily accept or recognize that dirty data can impact the health of companies..
  • 7.
  • 8.
    Building a CleanData Practice Establishing a Clean Data Practice is dependent upon a strong foundational Data Platform
  • 9.
    THE RIGHT REASONS CONSIDERATIONSFOR YOUR DATA PLATFORM
  • 10.
    Use Case Considerations ComplianceReporting Governed data produces certified results that ensure no miscues in both internal and external reporting Impact Analysis Change management can easily trace and identify any impacts to data consumers Digital Transformation Architecture should leverage a hub and spoke model to enable domain based micro service builds System Replacement Converting to a new system should leverage clean data as part of any data import activities Growth By Acquisition Requires a data strategy that supports a consolidated view of data across multiple data sources
  • 11.
    MAKING THE DISTINCTION SINGLESOURCE OF TRUTH VS BEST VERSION OF TRUTH
  • 12.
    Making the Distinction SingleSource of Truth Best Version of Truth Data storage principle to always source information from a single source Multiple sources of similar data across transactional systems Enables transparency, traceability, and clear ownership of the data Impacts timeliness and completeness of enterprise data Data usage principle for a single agreed upon view of data Requires a governed Master Data Management stewardship Results in certified “trusted” data for all data consumption needs Utilize business rules to eliminate data redundancy and define metrics
  • 13.
  • 14.
  • 15.
  • 16.
  • 17.
  • 18.
    OLAP Cubes Defining Characteristics •Daily data latency at minimum • Structured by analytical consumer functions • Semantic Layer with accompanying aggregation(s) • Data cubes enable consumers to quickly slice, dice, and summarize data in a presentation tool Typical Data Consumers • Production Support • Presentation Tools • Reporting Analysts • Executives / Upper Management
  • 19.
  • 21.
    Cloud Lake House Streaming Mobile LogFiles IoT Social On-Premises Databases Files Data Warehouse SaaS Applications ERP DATA SOURCES DATA GOVERNANCE Data Catalog | Master & Reference Data Management | Policies & Procedures DATA SECURITY User Provisioning | Protected Information | Network Access CLOUD DATA LAKE Raw Zone Structured Zone Curated Zone ANALYTICS SANDBOX Data Scientists CLOUD DATA WAREHOUSE Data Marts ODS OLAP Cubes CONSUMERS Data Analysts Presentation Tools Business Users APIs & Extracts CLOUD STORAGE STREAM PROCESSING BATCH PROCESSING
  • 22.
    Utilize the opportunityto hit the reset button Planning For Modernization Data Governance is critical to your success Avoid the pitfalls of a “lift and shift then fix” migration Start small with a focus to maximize data enrichment Take advantage of the ecosystem to avoid vendor lock
  • 23.
    Thanks For JoiningUs We hope you enjoyed the presentation. If you’d like to learn more about The Clean Data Initiative, we encourage you to download the full eBook. DOWNLOAD EBOOK www.sensecorp.com | marketing@sensecorp.com
  • 24.