Technology
Roundtable –
Business
Intelligence
and Analytics
© Solitaire Interglobal Ltd.
Kat Lind
Ms. K.R.E. Lind (Kat) is the Chief Systems Engineer at Solitaire Interglobal, Inc. (SIL). She has more than 45 years of experience in risk, analysis, general analytics and the management, design and implementation of large scale, high performance database systems. Kat is a frequent guest speaker at conferences and symposiums, spanning technical and user perspectives. She teaches a full curriculum relating to the design, implementation and tuning of database deployments at a graduate level. Ms. Lind’s expertise has been
acknowledged by published interviews in nationwide and international magazines, typified by a recent interview for IBM Systems Journal. Published extensively, Ms. Lind has authored articles, books on technical subjects and papers covering a wide range of topics.
Ms. Lind has been instrumental in developing SIL’s predictive performance modeling (PPM) which uses applied chaos theory and catastrophe mathematics. Under her direction, SIL has widened the scope of PPM beyond IT to areas such as marketing, general analytics, operational forex and more. Her in-depth, broad experience spans many industries such as finance, manufacturing, health care, government, transportation, etc. Ms. Lind’s technical expertise is considerable, as it has grown and evolved for more than 45 years of working with analytics, business intelligence, risk and large masses of data.
Objectives
• Provide a concise check list of critical
components and strategic decision points in each area of discussion
• Explore the experience of other
organizations to better understand the effects on cost, timeframe and other dimensions of deployment
• Meaningful dialog to address questions
Outcomes
• An understanding of the critical steps and
decisions that need to be addressed with business intelligence and analytics systems
• Insight needed to build a basic project
plan for business intelligence and analytics implementation
Business Intelligence and Analysis
All analytics shared issues and concerns, such as security, accessibility, risk and exposure. This talk will provide a view of
critical components and strategic decisions, backed by analyses of other organizations’ deployment experience in cost, timelines and risk that have already addressed some of the challenges that face FCCC today.
Understanding BI
• Different levels of analytics• Straight analysis, i.e., statistical
• Discovery of connections and patterns • Hybrid systems • Delivery • Canned reports • Ad hoc reports • Query drilldown • How BI is built
• Perception and understanding
• Correlation is not causation • Calibrated findings
• Actionable results
BI – Critical Components
• Shareholder sponsorship
• Functional group responsibility and buy-in
• Consensus on key ideas and the meaning of terms • Conceptual mapping to eliminate data confusion
• Establishment of a dynamic relationship between the BI view and the physical and logical data storage
• Setting expectations for living deliverables (growing and changing with time and experience)
• Creation of prepackaged BI objects that use the same data in different ways
BI by Design
• Business intelligence must be designed • Does not happen instantaneously
• A fundamental component is an understanding of the questions that will be
asked
• The users of the BI system have to be considered • Tool selection that includes targeted scope
• Form of analytics and modeling • Access mechanism, such as a portal
• Delivery vehicle, i.e., presentation, file extraction, etc. • Master data management
• Data integration
• Collaboration mechanism
BI and Extended Analytics
• Data visualization is a core and critical component
• Data discovery is inherent • Provides calibration data • BDA
• Provides additional discovery pathways • Tells data stories
• Uses data where it lives, using an organic model
BI and Integrated Analytics
• Integrated business intelligence model
confirms users are getting accurate, consistent information
• Assumptions, definitions, calculations, KPIs
and metrics are always originated from the same place no matter how used
Deploying BI
• Design and definition
• Initial construction of data collection
• Metric and monitoring skeleton creation
• Delivery mechanism implementation
• Evolutionary buildout, possibly in an agile
form
• Training and educational buildout
Deploying BI - Timelines
• Timelines are dependent on critical factors
• Scope of initial data • Variety of users
• Delivery mechanism
• Complexity of tools that will comprise the delivery
mechanism
• Dependencies
• Are not static, but change over time
Deploying BI - Risk
• Scope creep
• Poor understanding of capacity needed, both
people and machines
• Uncontrolled availability
• Security lax or non-existent
• Lack of training and support
• Static view adoption
Deploying BI – Critical Decisions
• Success in BI deployment rests on critical
decision points
• There are variances, but core ones remain
• Scope selection (data and users)
• Delivery mechanisms and components • Success metrics
• Monitoring method • Rollout plan
Case Study Comparison
Description Organization A Organization B
Data scope 85 tables, 500 TB 91 tables, 512 TB
Users 9 groups, 212 total 9 groups, 193 total
Functional levels 6 tools, 23 reports 7 tools, 20 reports
Planned deployment 6 months, 6 FTE 6 months, 5.5 FTE
Actual deployment 17.5 months, 14.5 FTE 6.5 months, 6 FTE
Adherence to
SIL Query Results
• Summary of over 1,400 similar projects within the last two years
• Examined for adherence to
guidelines, budget variance and
Metrics and General Comparisons
This was in 2008. The most current numbers shows that this is even higher with over 81% of BI software being “shelfware”.
Contacts
Kat Lind
Chief System Engineer Solitaire Interglobal Ltd. kat@sil-usa.com