What Can You Do with Machine Learning?

As machine learning has been around for decades, there have been many uses for this powerful technology. It also helps that there are clear benefits, in terms of cost savings, revenue opportunities, and risk monitoring.

To give a sense of the myriad applications, here’s a look at some examples:

• Predictive Maintenance: This monitors sensors to forecast when equipment may fail. This not only helps to reduce costs but also lessens downtime and boosts safety. In fact, companies like PrecisionHawk are actually using drones to collect data, which is much more efficient. The technology has proven quite effective for industries like energy, agriculture, and construction. Here’s what PrecisionHawk notes about its own drone-based predictive maintenance system: “One client tested the use of visual line of sight (VLOS) drones to inspect a cluster of 10 well pads in a three-mile radius. Our client determined that the use of drones reduced inspection costs by approximately 66%, from $80–$90 per well pad from traditional inspection methodology to $45–$60 per well pad using VLOS drone missions.”⁹
• Recruiting Employees: This can be a tedious process since many resumes are often varied. This means it is easy to pass over great candidates. But machine learning can help in a big way. Take a look at CareerBuilder, which has collected and analyzed more than 2.3 million jobs, 680 million unique profiles, 310 million unique resumes, 10 million job titles, 1.3 billion skills, and 2.5 million background checks to build Hello to Hire. It’s a platform that has leveraged machine learning to reduce the number of job applications—for a successful hire—to an average of 75. The industry average, on the other hand, is about 150.¹⁰ The system also automates the creation of job descriptions, which even takes into account nuances based on the industry and location!
• Customer Experience: Nowadays, customers want a personalized experience. They have become accustomed to this by using services like Amazon.com and Uber. With machine learning, a company can leverage its data to gain insight—learning about what really works. This is so important that it led Kroger to buy a company in the space, called 84.51°. It is definitely key that it has data on more than 60 million US households. Here’s a quick case study: For most of its stores, Kroger had bulk avocados, and only a few carried 4-packs. The conventional wisdom was that 4-packs had to be discounted because of the size disparity with the bulk items. But when applying machine learning analysis, this proved to be incorrect, as the 4-packs attracted new and different households like Millennials and ClickList shoppers. By expanding 4-packs across the chain, there was an overall increase in avocado sales.¹¹
• Finance: Machine learning can detect discrepancies, say with billing. But there is a new category of technology, called RPA (Robotic Process Automation), that can help with this (we’ll cover this topic in Chapter 5). It automates routine processes in order to help reduce errors. RPA also may use machine learning to detect abnormal or suspicious transactions.
• Customer Service: The past few years has seen the growth in chatbots, which use machine learning to automate interactions with customers. We’ll cover this in Chapter 6.
• Dating: Machine learning could help find your soul mate! Tinder, one of the largest dating apps, is using the technology to help improve the matches. For instance, it has a system that automatically labels more than 10 billion photos that are uploaded on a daily basis.

Figure 3-2 shows some of the applications for machine learning.