This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Close

If you need help, please click here:

Smart Analytics for Big Data

Big Data Analytics, Business Analytics, and Smart Analytics

Analytics is the discovery and communication of meaningful patterns in data by using statistics, computer algorithms, and operations research to quantify performance. Companies apply analytical techniques to business data in order to describe, predict, and improve business processes and performance.

Business Analytics applies statistical methods to business performance data to generate new insights for business planning by presenting the data in a visual form that is easy to understand.

Big Data Analytics uses specialized software tools and applications to analyze very large sets of data. Big Data Analytics helps enterprises to understand and identify the data that is most important for driving better business decisions in the future. Enterprises are always looking to find actionable insights into their data. With Big Data Analytics, an enterprise can find the answer to specific business questions, boost sales, increase efficiency, and improve operations, customer service, and risk management.

A smart solution that offers a single, integrated view of all information and business performance metrics is needed to add agility to decision making. With Smart Analytics, individuals in every function in any industry can leverage the results to improve business performance, as these examples show:

  • Executives can monitor the health of the enterprise in real time through dashboards, scorecards, and Key Performance Indicators (KPIs) for smarter, faster decisions and better business outcomes.
  • IT can use Smart Analytics to help speed business processes and business decisions, for example, by (1) determining when and where the customer experience breaks down, (2) discovering where inefficiencies lie in key business processes, (3) revealing when and where performance problems affect the bottom line, (4) uncovering where legal issues may arise (and how to prevent them), (5) finding the root causes that result in customer dissatisfaction, and (6) determining false indicators from past data.
  • Smart Analytics for IT can add value to algorithms such as (1) event correlation analytics, (2) topological relation analytics, (3) statistical pattern analytics, (4) textual pattern analytics, (5) configuration analytics, and (6) economic modeling analytics to optimize the usefulness of the data.
  • Marketing managers can perform better market basket analysis, improve segmentation analysis for targeted marketing, refine requirements for future products, develop offers and bundles for cross-selling and up-selling, and perform better campaign management by prospecting databases across all marketing channels.
  • Sales managers can engage in real-time sales analytics across all sales channels and territories, identify regional gaps, track product demand, and better match sales personnel to sales opportunities.
  • Financial managers can monitor and analyze all aspects of financial performance and risk in real time for budgeting, integration with R&D processes, and financial reporting and compliance, as well as identifying customer and supplier credit worthiness and fraud, and evaluating revenue, costs, and profitability.
  • Customer Care can work with a 360° view of customers to better serve their needs across all forms of communication and interaction, improving customer satisfaction and loyalty.
  • Operations managers can manage supply chains in real-time and better forecast inventories, more closely matching supply to demand.
  • Human Resource managers can more effectively balance resources with business and R&D needs, finding the right candidates for new openings.

With Big Data analytics, performance and capacity management functions can improve business operations across the entire enterprise. Big Data Analytics enables (1) correlating business processes with IT performance, (2) understanding how changes in business processes impact IT, (3) parsing IT costs by business unit/process, and (4) understanding business process performance at the server, storage, and network component level.

Software-Defined Data Centers

Software-Defined Data Centers (SDDCs) enable control over all aspects of the data center — computing, networking, and storage — through hardware-independent management and virtualization software. The SDDCs orchestrate, coordinate, and apply resources from the server, storage, and networking pools to ensure that the applications or services meet the capacity, availability, and response time that the business requires. This becomes possible when IT infrastructures evolve into more-scalable and manageable resources through the abstraction of servers, storage, networks, and applications, which makes these resources available to users as software-based services.

SDDC management models require real-time data collection, embedded analytics, and the ability to span multiple data source domains intelligently. The analytics can then meet different goals for increasing efficiency, reducing cost, and performing root cause analysis. Goals may also include workload-centric optimizations, global cost reduction, and improvement of energy efficiency along with global availability.

Thus, Smart Analytics become an essential component for reporting all kinds of knowledge within the SDDCs.

Smart R&D Cloud-Based Platform

A smart R&D platform based on private cloud infrastructure has a virtualized environment similar to the infrastructure and services available in a SDDC for cloud management, interface, elasticity, and scalability. A smart R&D platform enables sales, marketing, manufacturing, and other business centers to align seamlessly with R&D and IT processes for maintaining competitive advantage, spurring innovation, and developing better products. Smart Analytics is one of the essential components of cloud-based R&D platforms.

Software-Defined Smart Analytics for Software-Defined Data Centers and Smart R&D Platforms

The figure on the opposite page shows an R&D platform with Smart Analytics and software-defined servers, networking, and storage.

Software-defined Smart Analytics can analyze and generate all types of insights and reports based on complex analyses of combinations of IT and business processes. SDDC or R&D cloud platforms consist of a set of virtual or physical servers where applications draw resources from a heterogeneous collection of computing, networking, and storage resources. The ultimate goal of the SDDCs or R&D cloud platforms is to provide a set of software management tools that do not rely on proprietary solutions, making the initial setup and on-going optimization of Big Data application environments and continuous process improvement environments seamless for Smart Analytics.

With a collection of software-defined computing resources, software-defined network, Software-Defined Storage (SDS), a software-defined Hadoop MapReduce system, and other software-defined resources, organizations can define, manage, and optimize software-defined Smart Analytics. From a business perspective, transitioning to SDDCs or R&D cloud-based platforms increases resource sharing and security while providing better business alignment of the IT infrastructure and rapid provision of applications for enhanced agility. In this process, the SDDC or R&D cloud platform becomes more than a technology transition — it is a comprehensive paradigm shift away from a purely technology-centric approach to enterprise IT, towards truly delivering business solutions and values. Software-defined Smart Analytics enables such a paradigm shift.

Big Data is forcing enterprises to adapt storage methodologies beyond traditional file and block storage. Object stores and SDS mechanisms, in particular, are quickly gaining footholds in organizations that need to store massive amounts of unstructured data for Big Data Analytics. Some vendors are promoting object stores with new “data-defined” storage techniques. The concept of SDS aims at making storage “application-aware” and enabling server administrators, application managers, and even developers to request and provision storage in a policy-driven, self-service manner. Once provisioned, storage can be monitored and managed within its application context according to SLAs and performance requirements. Software-defined Smart Analytics can then exploit these “data-defined” storage facilities for machine learning and Big Data Analytics.

Software-defined Smart Analytics can leverage more application-centric and more automated networking infrastructure by using Software-Defined Networking (SDN) and network virtualization. Network virtualization is an evolutionary step from intra-hypervisor virtual switching and multi-hypervisor distributed switching. Virtual network overlays are programmable, offering APIs that can be connected into orchestration and automation platforms for Big Data transfers, transformations, and analytics.

The integration of software-defined Smart Analytics with enterprise processes running within the corporate SDDC or R&D cloud-based platform presents a significant challenge to enterprises. Software-defined Smart Analytics must include application and data storage patterns within their code and specifications for different marketing, sales, manufacturing, and IT requirements in order to run the application in a compliant and cost-effective manner. These patterns can also help systems and storage management software to automatically initiate troubleshooting measures as, for example, in the case of performance deterioration over Big Data storage systems, over a network during Big Data transfer, or over a security breach.

By Shyam Sundar Sarkar

Ph.D., Principal Architect, Corporate Strategy and Planning, Huawei R&D