Transcript Architecturing Business Intelligence For SMEs

BI
Business Intelligence
System
September 2013
www.bilisim.com.tr
Making Fact-based Desicions
Volume and complexity of data
flowing into companies increasing
Desicion-makers on all levels of
management with different job
roles need to understand this data
Yet, their time for decision-making
is decreasing
Challenges
More Data
More Users
Less Time
www.bilisim.com.tr
Implementing Business
Intelligence in the Companies
Challenges
 The organization hierarchy may not be suitable for a
new BI implementation project.
 IT departments may not have the required BI and
data warehouse skills.
 Limited IT budget.
 Different OLTP systems from different vendors with
little or no integration between them.
 Unstructured data which are stored in e-mails, a high
number of Excel Sheets and unconsolidated reports
and query results from a number of OLTP systems
www.bilisim.com.tr
Business Intelligence Applications
PRESENTATION LAYER
OLAP Tools, Dashboards, Reports,
Interactive Online Queries
Analysis
DATA WAREHOUSE / DATA MART
Metadata Repository, OLAP cubes
Data Loading
STAGING AREA
Data storing, cleansing, selecting, transforming, aggregating, integration operations
Data Extraction
DATA SOURCES
Online Databases (OLTP), Semi-structured data like Excel files.
www.bilisim.com.tr
The main components of a BI system
Decisions
Optimization
Data Mining
Data Warehouse / Data Mart
Data Sources
Business intelligence: set of tools and technologies that systematically exploit
available data to retrieve information useful in supporting complex decision
making processes
www.bilisim.com.tr
bilişimBI’s Approach
 Agile business intelligence
 Simple and intuitive user interface.
 No “high-tech terminology”, no “data warehouse
knowledge” needed to create multi-dimensional
analysis.
 interactive dashboard-style communication.
 All ETL (extract, transform, load) functionality
included. No additional license fees to be paid to
third-party software.
 Speedy response with in-memory query execution.
 The bilişimBI works seamlessly with bilişimERP.
www.bilisim.com.tr
The Architecture of bilişimBI Solution
Decisions
Optimization
Data Mining
Visual Query Tool
(Real time data analysis)
Data Exploration
Multidimensional
Analysis
Tool
Data Warehouse/Data Marts
META DATA
REPOSITORY
Operational
(OLTP) Data
ETL Wizard
Data Sources
Personal semi-structured data
bilişimERP
www.bilisim.com.tr
3 Staged Research & Development Project
Stage 3
Advanced Analytics &
Data Mining
Visualization.
Stage 2
DataStrategic
Warehouse, DataDecision Support
Marts, OLAP Cubes
Applications
Data Integration
Stage 1
Fundamental Data Warehouse
& Visualization Functions
www.bilisim.com.tr
STRATEGIC DECISION SUPPORT APPLICATIONS
Stage 2
Performance
Monitoring, Balanced
Scorecards
Descriptive Analytics
Analytical Workflows
Time Series Analysis
Dashboards, widgets
Mobile BI
Office Integration
Will be ready by the end of 2014
www.bilisim.com.tr
Case Study
9000
8000
TOTAL COST ($)
7000
6000
5000
4000
3000
2000
1000
0
Product A
Product B
Product C
PRODUCTS
Product D
www.bilisim.com.tr
Case Study
The Analysis of Product Costs
TOTAL COST ($)
9000
8000
TOTAL COST ($)
7000
6000
Internal
Operations
6000
5000
5000
4000
4000
3000
External
Operations
2000
3000
1000
0
Product A
Product B
Product C
PRODUCTS
By clicking on
product C…..
Product D
Materials
2000
1000
COST TYPE
www.bilisim.com.tr
Case Study
To achieve the required analysis:
1. Build the metadata. (A prerequisite of all the analysis tasks)
2. Create the domain of analysis
3. Specify the data tables of the domain.
4. Determine the measures.
5. Determine the criteria.
6. Save the analysis domain and schedule the ETL tasks.
7. Configure the analysis widget (Graph, table, etc.)
8. Use the interactive dashboard to see the results.
www.bilisim.com.tr
Building the meta data
Case Study
www.bilisim.com.tr
Creating the Analysis Domain
You can create new analysis (ETL
tasks) by using the Add button.
You can update the existing
Analysis simply by clicking on it.
Case Study
www.bilisim.com.tr
What would you like to analyze?
www.bilisim.com.tr
What would you like to measure?
Probable facts are listed automatically.
www.bilisim.com.tr
What are your analysis criteria?
www.bilisim.com.tr
Save and Schedule
www.bilisim.com.tr
Defining Data Quality Rules
www.bilisim.com.tr
Configure the dashboard settings
www.bilisim.com.tr
Interactive Dashboard
9000
9000
8000
8000
7000
6000
Internal
Operations
5000
4000
3000
6000
5000
4000
3000
2000
2000
1000
1000
0
TOTAL COST ($)
TOTAL COST ($)
7000
0
Product A
Product B
Product C
PRODUCTS
Product D
Product A
Product B
Product C
PRODUCTS
Product D
External
Operations
Materials
www.bilisim.com.tr
Collaboration with the Industry
Benefits for the Software Developers
✔ Ability to see real-life analysis requirements
✔ High-quality OLTP data to work with
✔ Useful feedback in the early phases of the project
✔ Extensive testing
✔ Ability to verify the design desicions
✔ Motivated agile development with short iterations
✔ Guidence about the functionality of future releases
www.bilisim.com.tr
Collaboration with the Industry
Benefits for the Industry
• Leveraged the value of their online enterprise data by
applying data quality rules
• Eliminated most of their heavy Excel work
• Leveraged real-time insights in the context of their
industry and role by using our Visual Query Tool.
• They even created “what-if” simulations to fully
understand the business impact of their desicions.
www.bilisim.com.tr
www.bilisim.com.tr