Big Data LDN 2018: FROM PROLIFERATION TO PRODUCTIVITY: MACHINE LEARNING DATA CATALOGUES
- 1. From Proliferation to Productivity with a Machine Learning Data Catalog Stephanie McReynolds, Vice President of Marketing @slangenfeld
- 2. Data is proliferating “data creation will swell to a total of 163 zettabytes by 2025” Enterprises are driving this “enterprises will create 60% of this data” Data consumers are growing even faster • Tableau Users 2018 = 78,000+ • Tableau Users 2014 = 26,000+Source: IDC Thought Leadership Practice Case Study, Data Age 2025
- 3. “90% of the time that knowledge workers spend in creating new reports is recreating information that already exists.” Source: KMM World
- 4. The social challenge of self-service analytics • Human data discovery is still in its infancy - Self-service analytics adoption has increased access to analysis - Understanding comes next and then data-driven decisions • We still rely on experts to Find, Understand & Trust data - Data distributed across 100s – 1,000s of sources - Nuances of data are often not documented - Self-service data prep is coming but early • Decision-maker understanding of visualized analytics is often: - Incomplete - Un-actionable - Inaccurate As a result, data-driven decisions are illusive.
- 5. "By 2020, organizations that offer users access to a curated catalog of internal and external data will realize 2x the business value from analytics investments than those that do not." - Gartner Magic Quadrant for Business Intelligence and Analytics Platforms, 2017
- 6. UseFind & Understand Trust Technical Metadata Business Metadata Social Metadata Data trains Machines guess Humans confirm Machines scale the learnings Feedback loop !! Machine Learning Data Catalogs
- 7. Data catalogs deliver break-through drugs to market faster “Data science shouldn’t be confined to mathematicians.” - JEFF KEISLING CHIEF INFORMATION OFFICER
- 8. The Pfizer data landscape was fragmented • 200B+ data Points • 180+ countries • 1M+ data sources • 500k+ tables • 6M+ columns • 6B+ queries “Data has become a very powerful currency for any company, and what we were finding is that the data was very fragmented.” - JULIE SCHIFFMAN VP OF BUSINESS ANALYTICS
- 9. The data catalogs linked queries & users by usage
- 10. Break-through ideas now come down to simple search…
- 11. …collaboration..
- 12. …and a diverse team of data literate employees.
- 13. Business outcomes driven by an Analytics Workbench • Goal: Deliver break-through drugs to market faster • Initiative: Identify rare disease markers, like transthyretin cardiomyopathy heart failures • Often goes undiagnosed because symptoms are similar to more common forms of heart failure • Analytics Workbench used to identify potential candidates • ML algorithms help with diagnosis and identifying clinical trial participants • Solution: Physician education + Targeted clinical trials Alation + Dataiku + Tableau
- 14. And more…
- 15. Data Catalog Outcomes to Consider Analyst Productivity Agile Stewardship Infonomics Increase analytic productivity 20-40% for data consumers. Accurate documentation up to 40% faster with machine learning. An interface to quantify the business value of all data assets stored.
- 16. For more information visit us in Booth # 473