Real-time Analytics Use Cases

Streaming analytics has emerged from being a domain-specific capability to having broad appeal across a range of industries and an increasingly diverse set of scenarios. Each use case benefits from the ability to derive real-time insights from large pools of data, leveraging the advances in sensors and smart devices, gateway technologies, in-memory storage, and new data management capabilities, as well as massive increases in networking and server resources.

At Intel, we are in a unique position to evaluate and learn from how streaming analytics can deliver unprecedented value to stakeholders, as the following examples show:

  • In clinical healthcare, real-time analytics provides the ability to monitor patient safety, personalize patient results, and assess clinical risk as well as reduce patient readmission—all of which improve organizational efficiency and the patient experience.
  • In transportation and fleet operations, trucking companies can use streaming data to pre-empt engine failures, reducing the risk (and considerable expense) of making repairs on the road. Traffic data can also be fed into a centralized dashboard, or returned to vehicles for driver action.
  • In retail and customer service, streaming analytics can improve operational efficiency, driving higher sales volumes, more trend insights, and enhanced customer satisfaction. And predictive models can help increase staff retention and reduce the recruitment overheads.
  • In finance, streaming analytics can aggregate data to improve business health and reveal insights about customer behaviors. Previous transactional data can be used to train machine learning models, to predict fraudulent transactions, enabling faster reaction times and reducing exposure to risk.
  • In manufacturing and supply chain, the low-hanging fruit is predictive maintenance, enabling servicing with minimal disruption and cost. Beyond the factory, supply chains for spare parts and raw materials can be optimized, as can forward capacity planning, process improvement and quality management.
  • In homes and small businesses, smart meters can be used not only to manage energy efficiency in the home, but also to feed data to secondary substations, so energy is available to meet demand. Predictive models can also be used to drive efficiency and reduce risk in small businesses.
  • In IT, systems and management data can enable better responsiveness, more timely capacity planning, and can feed continuous delivery and automated processes, driving the ability to manage different aspects of IT efficiently across infrastructure, operations and resources, and into business processes.

With such a bewildering range of possibilities, how can you decide which use of streaming analytics would be best for your organization? All of these scenarios share that it’s about the insight, so here are our top tips:

  1. Start with the business in mind. In another blog, we set out how to look for areas where the benefits of streaming analytics can be felt the most—these enable you to build credibility alongside growing skills. Note, however, that IT may have fixed ideas about what the business can achieve, which are not the most significant areas of potential. So, start by educating lines of business about what they might be able to achieve, then work with them on identifying the highest-value targets while incorporating the right levels of governance.
  2. Look outside the norms. At the same time, you can recognize that streaming analytics may enable you to do things that were not previously possible. Many innovative startups, in particular platform-based businesses such as Uber* or AirBnB*, have emerged because they operate in a space between existing business models. Ask yourself the question: what new business opportunities can be created based on the ability to know, or predict, what is actually going on in your industry?
  3. Focus on architecture. Streaming analytics success needs to minimize latency, deal with integration and burst requirements, depending on the scenario. For example the sensor-based data ingestion required for IoT scenarios will differ from the more transactional data needs of finance and customer service. You can maximize your chances of success by looking at best practice and architectural blueprints for the scenario concerned, deploying the right technology combination in a performance-optimized, efficient way.
  4. Work with experienced partners. One response, given so much choice, is to work with those who have experience deploying streaming analytics. At Intel we would be happy to help you understand the options, or recommend ecosystem organizations with experience in addressing needs such as yours. You can also research and review other areas of your industry, or industries with parallel challenges, to see how they are deriving benefits from streaming analytics.
  5. Recognize it is a journey. Streaming analytics is not a point-in-time solution, but a change in how we adopt technology to meet the needs of the insight-driven business. “We will start using prediction to perform tasks where we previously didn’t. Second, the value of other things that complement prediction will rise,” wrote Professor Ajay Agrawal and his colleagues, from University of Toronto’s Rotman School of Management. Our usage will evolve as both capabilities improve and our understanding of the benefits increases.

Streaming analytics should be seen not as an end in itself but as an enabler to new business success. Decision makers in today’s organizations are not familiar with having insights at their fingertips, and will need to learn how to adapt, to make the most of them. It is hugely important therefore, to ensure the business is put first in any streaming analytics decisions, both now and moving into the future.

To understand where to start on the journey to streaming analytics, download the whitepaper.