Big Data, Big Innovations

  • View

  • Download

Embed Size (px)


Think big data, and think opportunity. That is, think beyond storing and managing data, and leverage analytics to derive more value than imaginable from your business intelligence. This white paper offers a forward thinking, collaborative approach to analyzing data and changing the way you think about business.


  • 1.BIG DATA,BIG INNOVATIONSCOLLABORATIVE, SELF-SERVICE ANALYTICSDELIVERS UNPRECEDENTED VALUEForward-looking enterprises know theres more to big data than storing andmanaging large volumes of information. Big data presents an opportunity toleverage analytics and experiment with all available data to derive value neverbefore possible with traditional business intelligence and data warehouseplatforms. Through a modern, big data platform that facilitates self-service andcollaborative analytics across all data, organizations become more agile andare able to innovate in new ways.One question that enterprises are often faced with is, Now that I havebig data, what do I do with it? says Mike Maxey, senior director of strategyand corporate development at EMC Greenplum. Gaining value from big datarequires a new set of tools and processes that promote self-service dataexploration and analysis, along with social features embedded into every stepof the analytical process. Only through collaboration and knowledge sharingaround data sets, predictive model building, results, and best practices candata science teams evolve to uncover new insights from big data.Traditional data warehouse platforms break downWith a traditional data warehouse powering big data, its not unusual fordata loads and complex queries to run for days, which hinders the analyticalprocess. Plus, these data warehouse environments are often designed toanalyze structured data only, and not valuable unstructured data generatedfrom new external sources such as social media and mobile computing. Forexample, Twitter sentiment, GPS location data, web logs, sensor information,

2. WHITE PAPER | BIG DATA, BIG INNOVATIONS 2and many other forms of unstructured data all add to thegrowing body of data that can be used to better understandcustomers, manage risk, optimize operations and innovate.Big data requires a modern platform that is optimizedfor high-performance analytics across both structured andunstructured data. A query that takes 24 hours on a traditionaldata warehouse will take seconds on a modern analytics plat-form. Moreover, a modern analytics platform will have built-inadvanced analytics and data mining services, removing thelengthy process of copying data from a data warehouse intoa specialized analytics database or desktop application. Asa result, a modern analytics platform delivers more accurateand timely insight to decision makers who have a direct impact Havas Digital Helps Companies Increase Salesto the business. With EMC Big DataAt information security company McAfee, providingHear how EMC big data fundamentally changes the way Havas Digitalcustomers with timely information on potential security threats is analyzes data to provide better attribution for their clients marketingessential to the business. McAfees Security-as-a-Service emailefforts. web-filtering solution, powered by EMC Greenplum, relieson capturing and analyzing huge quantities of data and making itavailable to customers in a timely manner. Before using this plat- analysts are ready to perform the analysis, the supplied dataform, it would take customer service representatives upwards ofis outdated. Additionally, analysts experimenting with data and30 minutes to be able to answer a customers inquiry to find out building predictive models typically work in isolation with frag-what happened to a certain email. Now, with EMC Greenplum,mented data sets, without tools that centralize and documentwe have the ability to provide that service to a customer directly;insights to facilitate collaboration and knowledge sharing withthey no longer have to call in. They can go to our web portalpeers, IT and the business. As a result, predictive models areand identify within seconds what happened to [the email] as it not fully optimized and valuable insight is hidden and cannotwas flowing through our system, says Keaton Adams, principalbe reused across the organization.enterprise data engineer at McAfee. This not only improves If someone in marketing built a predictive model for productcustomer service for McAfee clients, but also allows the providerrecommendations, this insight and knowledge most likely is notto automate tasks that used to require human intervention. shared with, lets say, operations, which could leverage these best practices and methodologies to build a predictive model toThe advent of agile analyticsdetect fraud on the network, EMCs Maxey says.Big data also demands an approach to analytics that is flexible, To maximize the value of big data and impact business,accessible and fast. Traditional business intelligence tools provide analysts need a productivity tool to quickly provision their ownsophisticated analytics and data mining capabilities; however, sandboxes, and use their tool of choice to perform analysis,these tools tend to be rigid and hinder the analytical process.collaborate and share insights, and iterate on the entire With traditional tools, analysts must request from IT the process continuously.desired data sets needed to answer a question, and IT must Havas Digital Global, a media planning and buying company,create a reporting environment in which the analysis can berelies on EMC Greenplum to deliver innovative products andperformeda long and inflexible procedure. By the time the services in a highly competitive industry. Being able to quickly Groups of data scientists in multiple countries are able to come togetheron the development of Artemis, Havas Digitals analytics platform built onEMC Greenplum. This enables faster provisioning of sandboxes and bettercollaboration around testing and refinement of analysis and new analytics. Katrin Ribant, EVP, Data Platforms, Havas Digital Global 3. WHITE PAPER | BIG DATA, BIG INNOVATIONS 3 Greenplum Unified Analytics Platform (UAP) Greenplum Unified Analytics Platform combines the co-processing of structured and unstructured data with a productivity engine that enables collaboration among your data science team.experiment with data and collaborate around results allowsservices with EMC Greenplum.Havas Digital Global to better optimize marketing efforts Because big data is an emerging field, there is a shortage ofacross the web. data scientists. Fortunately, there are several education courses Groups of data scientists in multiple countries are able to offered by big data technology vendors and academic institu-come together on the development of Artemis, Havas Digitalstions to help expand the pool of available data scientists. Whatsanalytics platform built on EMC Greenplum. This enables fastermore impactful is an analytics platform that facilitates growth ofprovisioning of sandboxes and better collaboration around the data scientist community through knowledge sharing andtesting and refinement of analysis and new analytics, says collaboration that will ultimately bridge the talent gap.Katrin Ribant, EVP for data platforms at Havas Digital Global.EMCs answer to big data analyticsDisruptive data science With the right platforms, tools and expertise in place, enterprisesWhile having the right platform and tools in place to perform can develop big data analytics strategies that deliver results.big data analytics is important, so is securing the right people. EMC Greenplum has nearly a decade of experience in devel-Big data and the technologies that support it require a new oping products and services for data-driven organizations thatbreed of professionals called data scientists, who possess skills rely on high-performance analytics for business advantage. Bigand expertise that go beyond business intelligence. A datadata analytics is simply an extension to what EMC Greenplumscientist has a diverse skill set that combines statistical knowl-has always focused on, integrating new technologies aroundedge and programming expertise with business acumen and data science and Apache Hadoop to address big data analyticscommunication capabilities. But most important, a data scien- challenges so you can unlock insight from all your data sources.tist is a change agent with a passion for working with differentstakeholders across the organization to solve problemsmade Greenplum Unified Analyticspossible by gaining insights from new data sources. Platform (UAP)The data scientist is someone who can take raw forms Greenplum UAP provides the first and only modern analyticsof structured and unstructured data, internally and/or exter- platform that unifies all your data, with a productivity layer thatnally sourced, and apply advanced analytical techniques tofacilitates self-service and collaboration among data scienceuncover monetize-able and operationalize-able insights,teams. Through the integration of the following three compo-says Annika Jimenez, senior director of data sciencenents, data scientists can quickly start experimenting with bothA major part of big data analytics is getting teams productive early in theprocess followed by rapid iterations. Greenplum delivers freedom of tool selec-tion combined with Chorus for collaboration to enable rapid analytics at scale. Josh Klahr, VP of product management, EMC Greenplum 4. WHITE PAPER | BIG DATA, BIG INNOVATIONS 4structured and unstructured data, linking the data sets togetherto find insights never before possible. Greenplum Database for structured data an MPP database built for large-scale analytics processing and data loading, allowing for the consolidation and analysis of structured data stored in relational databases. Greenplum HD for unstructured dataprovides a complete and enterprise-ready version of Apache Hadoop, resulting in faster deployment, management and analysis of unstructured data. Greenplum Chorus for agilitya productivity layer that The Greenplum Unified Analytics Platform streamlines and centralizes the analytics process. DataBig data demands an approach to analytics that is flexible, accessible science teams are able to