In the last video, I explained the final two parts of the ML definition, predictive insights and repeat decisions. I also encourage you to think about the use cases within your business that really bring the core ingredients together. You might not realize it, but there's ML all around you and it's already transformed the way we're all living. Let's look at some examples from Google. Nest is a digital device that learns about your temperature preferences from the moment you turn it on. Have you used a voice assistant to call a family member? Google Home or assistant uses machine learning to recognize your voice and adapt to your favorite commands. You may have already seen predictive email responses in products such as Gmail. They use machine learning to predict phrases, you're most likely to increase the speed and efficiency of your emails. Google Photos uses machine learning to classify images. Maybe you've used a search within Google Photos to find the best picture of your family member or your dog, that's Machine Learning doing its job. Youtube's recommendation engine serves videos that would match your preferences based on what you've historically watched. You may have even noticed that many of these products use multiple ML models. This chart, for example, tells the story of Google's use of ML models over time. We weren't always using ML in so many of our products, but notice how the graph moves up over time. That is evidence that the number of Machine Learning models in use at Google has grown dramatically over the past few years, and it's not just high-tech companies that have seen their businesses transform by Machine Learning, this dramatic growth is now happening in many businesses around the world. One of the largest real-time auction service providers in the world, the process is more than four million auctions each year at some major pain points in their business processes. One major area of friction was image classification. To start an auction for one car required uploading 20 photos from various angles. It took a lot of time, up to 20 minutes per car, but the company had an idea, they built a real-time car image recognition system powered by Google Cloud's ML technologies. They shorten the amount of time it took to list the car auction from 20 minutes to only two to three minutes. Think about your business processes. Where do you have major pain points? How might ML help? As the cost of connected devices drops, unimaginable use cases open up. The Amazonian rainforest is rich with natural resources and cultural heritage, but it's threatened by people who illegally cut down trees. From a policy perspective, it's not possible to patrol that much land. But what if the marginal cost of monitoring was nothing? One company uses machine learning and small distributed listening devices to detect when illegal deforestation is taking place and contacts the authorities. Now, think back to the last time you ordered a pizza, maybe you had to pick up the phone to place your order. Well, Domino's, popular pizza delivery company has transformed the way their customers can order a pizza. They deliver over 10 million pizzas each week in the US. For years, customers have been able to order pizza via simple ordering bought available online and through the company's mobile apps. But they were determined to keep innovating. They wanted to build a better way for their users to interact with the bot to place orders. Using Google's dialogue flow chatbot technology, Domino's was able to improve the ordering experience even more for their customers. The bot could now handle more difficult or complex orders from customers. It's also provided a more natural and positive interaction for customers using the bot to place orders. With dialogue flow, they found ways to continue to improve their core business operations. Here are a dozen more ideas for using chatbots in different industries. You can use chatbots to resolve customer queries, for instance, or get travel advice or for credit applications, these are just a few of the many possibilities. Other businesses are using ML to re-imagine the way their users interact with products. Instead of mouse and keyboard, more and more people now use voice. In 2019, less than 50 percent of the internet users used Voice assistance. By the end of 2020, we can expect that more than 85 percent of customer interactions will be managed without human customer service representative, and by 2021, we expect that 87 percent of consumer interactions will involve chatbots or virtual assistance. Remember, data is an essential ingredient for ML, where there is data, there are ML opportunities. Take banking for example, there's so much data collected in banking. Think about some of the questions you might answer with all of the data available in banking, like, is this credit card transaction fraudulent? Should I offer this customer a loan or a savings account? How much will this current customer deposit over the next 10 years? Why is this customer calling the bank now? All of the ML examples we covered in this video use standard algorithms and involve volumes of data, predictions and repeated decisions. I hope they've helped you uncover even more ideas for your own ML's use case.
