Conversations between humans and machines have been a recurrent theme in science fiction. It is a reality now. Millions of people around the world are talking to Amazon’s Alexa, Apple’s Siri, Microsoft’s Cortana or Google’s Home. Technology is allowing us to use natural ways to interact with machines using touch, speech and gestures. We do not have to navigate complex menus or remember awkward commands to bridge the gap between intent and execution. Behind our interaction with machines, advances in speech recognition work with fast networks, sensors that build context, decision trees, behavior models and cloud-based analytical systems to decipher intent and desired outcomes. In fact, the more advanced the technology, the more it will fade into the background, becoming completely invisible, non-intrusive and natural.
Amazon Go[i], the groundbreaking 1,800 sq ft of retail space in Seattle, represents the latest in smart applications. Shoppers at the store check in using the Amazon Go app, pick the items they want and walk out. The store’s system knows which items were picked, adds them to a virtual cart, and charges the shopper’s Amazon account, instantly sending a receipt to the app. Amazon Go uses Deep Learning Algorithms, Computer Vision, Artificial Intelligence and Sensor Fusion to make this customer experience possible. Similar technologies are at work in autonomous cars, enabling simple and natural experiences. In the background, silently running the different technologies in perfect concert are smart platforms.
The ability to interpret complex human behavior is a special ability of smart applications. This is enabled through sophisticated learning processes. For example, when a shopper picks an item and returns it to the shelf at Amazon Go, the store system must understand precisely what was returned and delete it from the virtual cart. When an erratically moving object appears before an autonomous car, the system must be able to dynamically evaluate the threat it represents and take evasive action mapped to local laws. These actions call for smart processes to support the application.
Finally, smart applications interact with innumerable internal and external systems, devices, sensors, databases and networks to gather and exchange data, analyze information, and to determine and deliver action. User information can easily be exposed at various touchpoints between systems. Securing user data without compromising user experience can be tricky. These smart applications therefore demand smart security or adaptive security architecture that responds to dynamic threat environments using a continuous AI-driven feedback loop.
Smart applications are remarkable. They are simple on the outside and smart on the inside. They provide us natural ways to exchange information with machines, letting them get to know us, learn context, rely on AI to take decisions and then act proactively on our behalf, completely humanizing the experience. The question is, “What will it take to humanize your business?”