Enterprise Data Unification Powered by Machine Learning Jerome - PowerPoint PPT Presentation

Enterprise Data Unification Powered by Machine Learning Jerome Gransac Confidential Confidential

HI-IS Big Data Event Intro ● Tamr Overview ● ○ Background ○ Use Cases & Customer Success ○ Technology & Differentiation Product Demo ● Confidential Confidential

State of the Data in the Enterprise: Large Data Debt Reality : Constant change & entropy have created Requirement : Accurate, up-to-date, curated view of “Random Data Salad” core business entities Customers Products Transactions Suppliers Parts [...] M&A “Data Hoarding” Politics Leadership Changes Dynamic Schema DBs - Restructuring Legacy Mongo et al Burden Problem: 1. Too much time spent on data prep vs. analysis & action 2. High failure rate of BI & analytics projects 3. Game-changing initiatives deemed ‘impossible’ and never start Confidential

Tamr At A Glance Key Founders Company Overview Tamr solutions unify enterprise data by combining machine Dr. Michael Stonebraker learning and human expertise to power transformational analytic Co-Founder & CTO and operational outcomes. Previously: Founder, Ingres/Postgres, HP Vertica Headquarters: Cambridge, MA Andy Palmer Additional Offices: San Francisco, London Co-Founder & CEO Previously: Founder & CEO, HP Vertica Founded: 2013 Confidential

What Tamr Does: Enterprise Data Unification Tamr uses machine learning to attack the enterprise data variety problem to power transformative analytic and operational outcomes 10x Reduction $500M+ Savings Customer Insights 1,500+ Studies In new data set integration From sourcing analytics Unified buyer profiles across Unified clinical study data to from 6 months to 2 weeks across siloed business siloed dealer systems in 30+ geos empower researchers Video Case Study Video Case Study Video Case Study Case Study Spend Classification Inventory Optimization Product Sales Insights Well Productivity From pilot to live on Google $100M in reduced inventory by Unifying product sales data from Integrating disparate data related to Cloud Platform in 6 weeks harmonizing parts across 5 fleets distributors to enable new analytics wells to optimize productivity Video Case Study Confidential

Tamr Unify: Platform Overview Internal Data COMBINE CONSOLIDATE CLASSIFY BI / Analytics Schema Mapping Record Matching Classification Data Wrangling Custom Apps External Data Machine Learning Expert Input Microservices: RESTful APIs Automated Integration Source Remediation Confidential

Tamr’s Three Core Capabilities in Action Building Data Integration Probabilistic Models for Mastering & Classification Source1: Vendor_Name Unified Name Source2: Supplier Supplier Source3: LFA1-MANDT Source4: CompanyName SourceN: Nom_de_Cie Combine thousands of sources ● Extract signal from underlying data ● to drive mapping decisions Increase levels of mapping ● automation as sources increase Confidential

Tamr’s Three Core Capabilities in Action Building Data Integration Probabilistic Models for Mastering & Classification Husky Allied Husky = Manufacturing Incorporated Automatically group similar records or ● using machine learning 342 Suite 34 342 Main St KY KY, USA Infuse expert feedback easily and ● quickly for maximum accuracy Create ‘golden records’ ● Confidential

Tamr’s Three Core Capabilities in Action Building Data Integration Probabilistic Models for Mastering & Classification Should “Husky #.25 J Blt” in table “Invoices” be categorized as Automatically map records into any ● Hardware > taxonomy Fasteners > Leverage standard or custom n -tier ● Bolt classification scheme Easily adjust as needs evolve ● Confidential

Use Cases & Customer Success Confidential Confidential

Case Study: General Electric Agile, multi-domain entity mastering drives $500M+ in Value Technical Challenge Technical Outcome Suppliers : Build an integrated view from < 6 months from pilot to globally ● ● 75+ ERP systems and 2M supplier deployed data pipeline; 2M records records consolidated to 700k Parts: 25M non-unique parts in 25M reduced to 6.4M unique parts; ● ● purchasing systems across 8 BUs 5-tier analytic-ready classification ● M&A: Integrate data from acquired ● Data from 3 acquisitions integrated “ The supplier data integration was a big entities with existing master views with GE’s in <2 weeks win. ” Bill Ruh CEO GE Digital & Business Challenge Business Outcome Chief Digital Officer at GE “ We’ve seen firsthand the transformative Suppliers : Get GE’s best terms with a $80M savings in Year 1 from Tamr- ● ● results that Tamr’s technology has on an given supplier in every negotiation mastered unified supplier view enterprise of GE’s scale. When the cost and complexity of bringing together enterprise Parts: Optimize sourcing strategies to $300M in annual savings identified ● ● datasets is massively reduced, the resulting most cost-effective suppliers (0.5% reduction of direct spend) analytic breakthroughs create opportunities M&A: Increase the velocity of realizing Supplier, purchasing, and customer ● ● that were previously inaccessible. ” synergies post-acquisition base opportunities quickly identified Lisa Coca Managing Director, GE Ventures Confidential

Case Study: Thomson Reuters Optimizing an information company’s data curation operations Technical Challenge Technical Outcome ● Increase levels of automation in data ● Automation levels as high as 90% in processing key projects ● Capture and leverage expertise of data ● SME’s knowledge incorporated stewards Tamr’s models Improve data quality Sustainable precision and recall rates ● ● in excess of 95% “ Since we brought Tamr in four years ago, it has become an integral part of our big data platform which powers our products and services. ” Business Challenge Business Outcome Mona Vernon CIO Thomson Reuters Labs ● Accelerate time to market ● Months shaved off new product intro “ Tamr’s novel integration platform enabled Reduce manual effort 40% reduction of manual integration ● ● us to expedite our own entity integration Support increasing cloud adoption Hybrid on-prem / cloud deployments efforts by several months while reducing ● ● the manual effort by over 40% – a ● Take on data integration projects ● New TR products launched that were substantial achievement. ” previously deemed ‘too hard’ previously stuck on the drawing board Tim Baker Global Head of Content Initiatives Confidential

Case Study: Toyota Motor Europe Unified customer 360 views from highly fragmented data collection Technical Challenge Technical Outcome Effectively integrate locally managed ~125 sources currently integrated; 500 by ● ● customer data from 30 countries project end Maintain flexibility for data collection No disruption to local systems, but new ● ● and management at the country level mastered data now available Integrate new sources quickly 1 - 2 weeks on average ● ● Support local migration projects TME France CRM migration completed in ● ● “A lack of consistent view of customer data 6 weeks was restricting our ability to innovate and meet the expectations of our customers… Business Challenge Business Outcome Addressing these issues led us to an enterprise data unification approach, and a vendor, Tamr. We rejected traditional Deliver better service for customers who CSRs now have single UI to search ● ● commercial offerings like MDM tools move between countries Tamr-mastered customer data because their top-down approach required a single data model.” Provide consistent experience across all Better customer knowledge at point of ● ● customer touch points sale / service Matt Stevens Understand and predict the needs of First ever unified view of customers ● ● Director of Information Systems customers to improve ability to exceed fueling new analytical and operational expectations use cases Confidential

Case Study: GSK Unified R&D data lake from fragmented research domain silos Technical Challenge Technical Outcome R&D data too siloed / fragmented to be Used Tamr’s “probabilistic matching” to ● ● used effectively for exploratory combine data into a single data lake with purposes. three different domains. A traditional master data management ● ● All assay, clinical, and genetic data approach would have taken too much moved into lake and unified in 3-months time and effort to implement. “ GSK R&D’s data environment is something that one ofuen hears about in startups, but is rarely found in large enterprises whose Business Challenge Business Outcome roots go back over 300 years. And it’s great news for all of us humans who will benefit from the scientific advances it is likely to New drug development cycles impeded Tamr’s machine learning-based enterprise ● ● engender.” by inflexible, weakly integrated data data unification platform transformed the environment company’s data management capabilities. Tom Davenport Needed transformation in how data and Made it easier to access and use data for Biting The Data Management Bullet At GSK, FORBES ● ● analytics were used across the exploratory analysis and decision making organization about new medicines Confidential