Implementation of data science in organizations
0 likes525 views
The document outlines the implementation of data science within organizations, highlighting its definition as the use of data and analytics to enhance business performance. It describes three maturity stages in data science: data organization, business intelligence/reporting, and analytics implementation, emphasizing the importance of data quality, leadership, infrastructure, human resources, and strategy. Key areas for successful implementation include ensuring data accessibility, establishing a fact-based culture, and embedding analytics into business processes.
Implementation of data science in organizations
- 1. Implementation of Data Science in Organizations Koo Ping Shung koolanalytics@gmail.com
- 2. Agenda What is Data Science? Three Stages of Data Science Maturity Getting Data in Order Getting Biz Intel/Reporting Process in Order Let’s do Biz Analytics/Data Science Areas of Focus for Successful Implementation Data Leadership Infrastructure Human Resource Strategy Embedding into Processes
- 3. What is Data Science/Analytics?
- 4. 4 What is Analytics/Data Science? There is no formal definition now. Used to be called Analytics but have moved on to be called Data Science. Lack of formal definition have allowed for a wide interpretation.
- 5. 5 What is Analytics/Data Science? Analytics is the extensive use of data, statistical and quantitative analysis, explanatory and predictive models, and fact-based management to drive decisions and actions. ~ Thomas H. Davenport Competing on Analytics:The New Science of Winning
- 6. 6 What is Analytics/Data Science? Put it simply, It is IMPROVING Performance in KEY business domain using DATA and ANALYSIS
- 7. Three Stages of Data Science Maturity Getting Data in Order Getting Biz Intel/Reporting Process in Order Let’s do Biz Analytics/Data Science koolanalytics@gmail.com
- 8. Getting Data in Order Collecting the ‘Right’ Data Find the low hanging fruits. Establish Data Collection Process Existing Cheaper Data Management High Quality Data Sufficient Quantity koolanalytics@gmail.com
- 9. Biz Intel/Reporting Process According to Strategy and KPI. Day to Day Operations Data Needed ETL Process setup – Infrastructure (Cloud perhaps?) Data Storage koolanalytics@gmail.com
- 10. Let’s Do Data Science Explore current process How Data Science can improve on it. Labor Intensive or Capital Intensive (Costly!!) Collects a lot of Data Scalability and Consistency koolanalytics@gmail.com
- 11. Moving to Biz Analytics/Data Science Data – Accessibility and High Quality of Data Leadership – Analytical Leadership Infrastructure – Highly Integrated Infrastructure Human Resource – Attracting and RetainingTalents Strategy – Setting the Right Objectives & Performance Metric Embedding – Embedding Models into Processes & Operations koolanalytics@gmail.com
- 12. Data – Accessibility and High Quality Data Data Management Collection Channels – Pros and Cons Accuracy – Absolute or some Margin of Errors Timeliness – OnTime and Frequency of Collection Missing Data - Handling Data Errors – Investigation and Rectification Governance – Authority Types of Data – Structured and Unstructured Accessibility – Who can and at what level? Back Up – When and How? ETL Process – Risk and Errors koolanalytics@gmail.com
- 13. Leadership – Analytical Leadership Besides the usual leadership skills Establishing a Fact-Based Culture Autonomy Understand the Limits of Data Science Building an Ecosystem that Encourage Innovation and Experimentation. Change Management Conflict Management koolanalytics@gmail.com
- 14. Infrastructure – Highly Integrated Infrastructure Data Capturing/ManagementTools and Processes DataTransformationalTools & Processes Data Repositories Reporting,Analytical/ModelValidationTools PresentationTools & Processes Embedding Analytics Metadata Operational Processes koolanalytics@gmail.com
- 15. Human Resource – Attracting and Retaining Talents Build an environment that provides support and research. Team Players. Talents are hard to find. Know what makes them ‘tick’. Team Lead Change Management Conflict Management Establish Clear Boundaries and Authorities Ground Rules on Working Together koolanalytics@gmail.com
- 16. Strategy – Setting the Right Objectives & Performance Metric Actionable insights to Establish Right Strategy Understand Internal and External Environment Setting Right Performance Metrics Feedback loop if Strategy is working Allow for changes if it does not rather then looking at the final results. koolanalytics@gmail.com
- 17. Embedding – Embedding Models into Processes & Operations Gain value from Models Reduce Costs and gain Consistency Scalability Infrastructure Support Execution Challenges koolanalytics@gmail.com