Using Power BI and Azure as analytics engine for business applications
Formulating Power BI Enterprise Strategy
1. Formulating a Power BI Enterprise Strategy
Teo Lachev, Prologika
Brian Jackson, Microsoft
2. Agenda
• 8:30-9:00 – Networking and introductions
• 9:00-10:30 – How Power BI empowers businesses like yours
• 10:30-10:40 – Break
• 10:40-11:45 – Plan a data access strategy, hybrid architecture case study
• 11:45-12:00 – Q&A
• We welcome questions during the presentation
3. Power BI Workshop
2-day Applied Power BI Workshop
Sep 14-15, 8:30-5:00, Microsoft Office in Alpharetta
Reserve your seat today to attend this insightful 2-day
workshop for only $999
use coupon POWERBI20160914 to get 10% discount when
signing up two or more people)
bit.ly/powerbiworkshop
4. About Teo
• Consultant, author, and mentor in Data Analytics
• Owner of Prologika – consulting and training company
based in Atlanta (prologika.com)
• Microsoft Gold Partner in Data Analytics
• Power BI Red Carpet Partner
• Microsoft SQL Server MVP since 2004
• Founder and leader of Atlanta BI group (atlantabi.sqlpass.org)
5. About Brian
24 years of industry experience, 10 years at Microsoft
• Data Solution Architect, Microsoft
• Many Fortune 500 implementations of SSAS, SSRS, SSIS, Office 365, SharePoint
• Prior to Microsoft – HP, US Air Force
• Fun fact: started as an IBM 370 Mainframe Assembler Programmer
Passion for Certification
• Passed all Microsoft Azure exams: 70-532, 70-533, 70-534, 70-473, 70-475
• Microsoft Certified Solution Master Data Platform
• Helped write first SQL Server Business Intelligence exams
• Certified on every version of SQL since SQL 7.0 in 1998
6. What's Power BI?
• An umbrella name for:
• Power BI Service (powerbi.com) – a modern cloud-based platform for sharing BI content
(weekly updates)
• Power BI Desktop – free Windows app for self-service data modeling and building reports
(monthly updates)
• Power BI Mobile – a set of native apps for iOS, Android, and Windows devices
• Power BI Embedded – a cloud service for integrating Power BI reports in custom apps
• What Power BI is not
• Power BI for Office 365
• SharePoint-dependent
• Excel-dependent
• Replacement of SSRS
• Immature technology
9. For Business Users
• Connect to many cloud and on-premises data sources
• Import and analyze data from Excel files
• Analyze data from popular cloud services
• Instant report and dashboards
• Get predefined dashboards from content packs
• Create ad-hoc reports
• Ask questions through natural language query (Q&A)
• Share dashboards with your coworkers
• Mobile applications and alerts
• Native mobile applications for iPad, iPhone, Android, and Windows
• Create and receive alerts and important data changes
10. Power BI Mobile
https://powerbi.microsoft.com/mobile
• Download native apps for iPad, iPhone, Android, and Windows
• View and change reports
• Annotate and share
• Receive alerts about important changes (iPhone app only)
12. For Business Analysts
• Connect to all of your data
• On-premise databases and files
• Cloud-based solutions
• Clean, transform, and shape data
• Design sophisticated data models
• Create powerful business calculations
• Two options to create data models and publish to Power BI
• Freely-available Power BI Desktop
• Microsoft Excel
15. Classic Organizational BI Solution
Data Sources
Data is extracted from
data sources,
transformed, and
loaded into DW
Data Warehouse
Data is stored in
dimensional schema
consisting of dimension
and fact tables
Dimension
Tables
Fact
Tables
ETL
Integration Services
Semantic Layer
Great performance
Business calculations
Reduced reporting effort
Security
Descriptive & predictive analytics
Multidimensional
Tabular
Report Clients
Standard reporting
Ad-hoc reporting
Dashboards
Power BI
Excel
SharePoint
Third party Tools
Analytical
Reporting
16. For BI Pros
• Establish trustworthy environment
• Content sharing with workspaces
• Content sharing with content packs
• Set up gateways to access on-premises data sources
• Implement hybrid BI solutions
• Single version of truth
• No need to move data to the cloud
• Performance
• Security
Live Power BI
dashboards and reports
Cloud
On-
premises
SQL Server
Analysis
Services
Live
Query
20. 10 Criteria Where Power BI Excels Competition
Overall
1. A component of MS Data Platform – no tool is a kingdom on its own
2. Cloud first – agility, no installation footprint, integration
3. Cost – Power BI Desktop is free, Power BI Mobile is free, Power BI Service via O365 E5 plan
For Business Users
4. Easy to get started with content packs and get data
5. Productivity features – Q&A, Quick Insights
21. 10 Criteria Where Power BI Excels Competition
Continued
For Data Analysts
6. Data shaping and transformations – cleanse and shape data
7. Sophisticated data models – on a par with models built by BI pros
8. Powerful programming language – implement any business calculation
For BI Pros
9. Flexible data connectivity options – live connections, hybrid architectures
For Developers
10. Extensible platform
Power BI Embedded
Custom visuals
APIs for access to content, such as for pushing data (real-time dashboards)
23. Planning Data Strategy
Data Import Live Connection
Data location Original data source
Power BI cached dataset
Original data source
Data size Constrained by available memory on
desktop
Limited to 1 GB in Power BI Service
None
Report queries Send to data model Sent to original data source
Performance Excellent and predictable Depends on the data source
Data refresh Required to sync changes
(except when models are deployed to
OneDrive or SharePoint Online)
Not needed
Gateway Required for on-prem refreshable data
sources
Required for direct
connection to on-prem data
sources
Mash up data
from multiple data
sources
Yes No
Supported data
soures
Many
SQL Server, Oracle, Terradata, DB2, MySQL,
SharePoint List (on prem),
Files (CSV, XML, Text, Excel, Access)
Folder
Excel Power Pivot (via Power Query only)
Power BI Designer
Cloud
SQL Azure Database
Azure Data Warehouse
Azure Spark
Amazon Redshift
(soon) Impala
On premises
Analysis Services
SQL Server
SAP Hana
Oracle
Teradata
Power BI
Datasets
Cloud Services
(DirectQuery)
On-premise Data
Analysis
Services,
SQL Server,
SAP Hana,
...
Enterprise Gateway
(DirectQuery and
refresh)
Refreshable
Data Sources
(RDBMS, Power
Query, ...)
Personal
Gateway
Cloud Services
(no direct connect)
Liveconnection
Import/Refresh Service
24. Feature
Connect live to your existing on-premises
data, including SQL Server Analysis
Services
o You can keep your data on-premises and
enable BI analysts / end users to create
Power BI reports
o Connect to and refresh data securely via
Azure Service Bus (SSL encrypted
connection)
o Service Bus only uses an outbound port, it
does not require an inbound port to be
opened on the on-premises firewall
Live Power BI dashboards
and reports
Cloud
On-premises
SQL Server
Analysis
Services
Live
Query
SaaS solutions
e.g. Marketo,
Salesforce, GitHub,
Google Analytics
On-premises Data
Organizational
Corporate data sources
or external data
services
Azure services
Azure SQL, Stream
Analytics…
Connect live to your on-premises dataData sources
25. Feature
Realize the benefits of a SaaS-based BI
solution without moving your data to the
cloud
Everything in the service is encrypted -
Power BI uses HTTPS for all traffic
For on-premise sources, e.g. SSAS, role-
based and row-level data security is applied
Data remains on-premise, only query is
stored in Power BI
Power BI requires valid accounts that meet
the requirements of the specific data source
Please see table for authentication methods
Keep your data where it is and connect securely
Authentication
method
Select data sources
that use this method
Basic SQL Azure, Marketo, ODATA Feed
Account key Blob store. Table store, HDInsight
OAuth
Salesforce, Google Analytics,
ODATA Feed
Marketplace key Azure Marketplace
Anonymous ODATA Feed, Web
Data authentication methods
Data sources
SaaS solutions
e.g. Marketo,
Salesforce, GitHub,
Google Analytics
On-premises Data
Organizational
Corporate data sources
or external data
services
Azure services
Azure SQL, Stream
Analytics…
Data sources
26. Feature
o Azure SQL Database
o Azure SQL Database Auditing
o Azure Stream Analytics
o Azure SQL Data Warehouse
Consume structured/unstructured data from other
Azure services
HDInsight
Storage
Event
Hubs
Machine
Learning
SQL Server
Stream
Analytics
Power BI
Data sources
SaaS solutions
e.g. Marketo,
Salesforce, GitHub,
Google Analytics
On-premises Data
Organizational
Corporate data sources
or external data
services
Azure services
Azure SQL, Stream
Analytics…
Data sources
27. Feature
Expand with Azure Services and Power BI
Power BI
HDInsight
Storage
Event
Hubs
Machine Learning
SQL Server
Stream
Analytics
Data Factory
Data sources
SaaS solutions
e.g. Marketo,
Salesforce, GitHub,
Google Analytics
On-premises Data
Organizational
Corporate data sources
or external data
services
Azure services
Azure SQL, Stream
Analytics…
Data sources
28. Scenario 1
Data Analyst and Heterogeneous Data
• Data is imported into a single model
• Power BI Desktop
• Excel
• Model is deployed to Power BI Service (size must be <= 1 GB)
• Refresh is scheduled
• Enterprise or personal gateway must be installed
• Unless the model is saved to OneDrive or SharePoint Online
29. Scenario 2
IT Builds a On-prem Semantic Model
• SSAS semantic model is build on top of data warehouse/data mart
• Single version of truth
• Deployed to a dedicated server
• Scale to terabytes
• (Alternative) Direct access to RDMBS (SQL Server, Oracle…)
• Direct Query is used
• No data is deployed to Power BI
• No refresh is required
• Enterprise gateway is installed on premises
30. Scenario 3
IT and Direct Query Cloud Source
• Direct Query is used
• SQL Azure Database
• Azure Data Warehouse
• Amazon Redshift
• Azure Spark
• Impala
• No data size limitation and no refresh is needed
• No gateway is required
• Performance depends on data source
31. Scenario 4
IT and Non-direct Query Cloud Source, e.g. Salesforce
• Data is imported
• Content pack (up to 1 GB data size)
• Custom import in Power BI Desktop
• Solution template can be used (currently Salesforce, Dynamics CRM and Twitter
only)
• Gateway is required if data is staged to on-prem database
33. Thank you!
Please feel in survey
1. How did we do?
2. How can we improve?
3. What other public events you're interested in?
http://bit.ly/powerbiseminarsurvey
Stay in touch:
teo.lachev@prologika.com
@tlachev
http://prologika.com (blog, newsletter, sample code, …)