big data: where is the beef? drill through the hype to get business value hans-josef jeanrond vp...

22

Upload: abdiel-brinkworth

Post on 28-Mar-2015

214 views

Category:

Documents


0 download

TRANSCRIPT

  • Slide 1

Slide 2 Big Data: Where is the beef? Drill Through the Hype to get Business Value Hans-Josef Jeanrond VP Marketing, Sinequa Slide 3 Where is the Beef Slide 4 A Mighty Question increase in revenue for Wendys in 1984 the year of the Where is the Beef commercial with Clara Peller 31% Anticipate this question in your Big Data plans! Gary Hart loses the Democratic primaries against Walter Mondale who asks Where is the Beef looking at Harts enormous case files Slide 5 Where is the Beef in the Big Data Hype ? Where the Beef ? is Slide 6 Big Data - Hype and Useless Disputes Useless: What is Big Data If its big and varied enough to cause you problems in dealing with it and if it may be relevant for your business: take it as Big Data An urban legend! Invented from A to Z False promises: You will be Rich The story of the nappies and the beer The Eternal Promise: Structure the World to Possess it This is the Mediaeval Solution: The Grail Slide 7 Big Data - Hype and Useless Disputes Beware: your Grail-quest companions may be the wrong company for the new Big Data world! Peut-tre peut-on illustrer l'ide que les anciens compagnons de route ne sont plus les bons, par un chevalier face qq quip d'un smart-phone / d'un GPS / d'une voiture / d'une fuse, ou qq chose dans le genre. Slide 8 Big Data One Certainty Big Data aint Structured! For the Brave who dare to confront many different languages and know how to analyze them And who dont just look at the meta data of documents! 80% of enterprise data is text = unstructured = outside the grasp of enterprise applications (ERP, CRM, BI, etc...) = inside documents, emails, social Networks, etc. Slide 9 Where is the Beef in Big Data? Look at just 2 Use Cases: Creating 360 views What has Big Data got to do with this? Mapping Implicit Social Networks of Experts Slide 10 360 View of a Customer Shares management Operations Credits Credit cards Portfolio Insurances Contracts Emails Letters Interactions with Call Centers Contract models Offers (product datasheet) Credit ratings Info Dunn Slide 11 360 View: Why Big Data? Volume: Index billions of transactions and hundreds of millions of documents Variety: Index transactions, ERP and CRM data, contracts, product brochures, emails, etc. Velocity: Get it all done fast enough to present an up-to- date view of the customer walking up to you or calling you Real-time is key! Hadoop batch processing is not good enough Slide 12 Implicit Social Networks? Why? Find the key experts in a given domain to build successful project teams. Key experts can help capitalize on existing skills, technology, and developments optimizing R&D organizations, and shortening time to market. Enterprise Social Networks dont work: Self-declared expertise in user profiles of Enterprise Social Network is too often out of date, incomplete or exaggerated Reveal Implicit Social Networks of Experts How? Finding the true experts requires looking at their work: Analyzing publications, project reports, patent filings, HR data and schedules, Enterprise Social Media content, emails, etc. This analysis requires Natural Language Processing (NLP) capacities to understand what topics people have written about. Slide 13 Implicit Social Networks: Why Big Data? Analyze up to 500 M documents in different languages Research and project reports Patents Articles in specialized journals Emails Etc. Go though Internal and external databases Etc. Slide 14 Automatically link Skills to People -> Expert discovery in Pharma Index of tens of thousands of People With Skills = Experts Hundreds of millions of internal and external documents With Skills Drugs Diseases Genes Brands MOA Slide 15 Expert Discovery in Manufacturing Index of tens of thousands of People With Skills = Experts Hundreds of millions of internal and external documents With Skills Components Supply Chain Master Data Competitor Products Tools Slide 16 Expert Discovery in Finance Index of tens of thousands of People With Skills = Experts Hundreds of millions of internal and external documents With Skills Loans Credit Risk Credit Cards Central Bank Real Estate Slide 17 Now for the Beef 360 Customer View Call Center: 60 M / year Customer facing agents and managers in branch offices: >11,6 M in year 2; >13 M in year 3 Revealing Implicit Social Networks Siemens: one successful Search can save tens and even hundreds of thousands of Euros Atos: The benefit is beyond ROI we cant do without it. AstraZeneca: Ask Nick (Nick Brown) Slide 18 Concrete ROI Calculation Sand Dune ROI Not taking into account new business generated by the use of Search. Productivity Gains According to a Scenario of Generalised Search within an Enterprise (Source: IDC) Slide 19 How do you get at the beef? Slide 20 REAL TIME Search, Discovery & Visualize Structured and unstructured data 120 Connectors Unlike NOSQL Deep content analytics Unique Semantic & Natural Language Processing Engine Big Data Search Analytics Unlike Hadoop High Performance, Extreme Scalability A unique positioning in the Big Data ecosystem Slide 21 Sinequa - More Beef and more Fun REAL TIME Big Data AND More fun to work with Sinequa - the AND Company Analytics Search Slide 22 All the ANDs are necessary! Sinequa - the AND Company Dont buy a comb with missing teeth Slide 23 MERCI Hans-Josef Jeanrond VP Marketing, Sinequa