Decision intelligence

In her book Link, research scientist and entrepreneur Dr. Lorien Pratt introduced "decision intelligence" and explained how to use artificial intelligence to connect decision makers to data and models. Of course utilizing information to support decisions is an essential tool that all animals use -- think of the information we process with our eyes -- is that a friend or an enemy coming towards me? But the massive amounts of information now available exceed the ability of our unaided human senses. Machine learning models and data visualization can bring this information to us in usable ways and allow us to make better decisions at a far greater scale and speed.

Take for example a marketing problem - how do you send a personalized message to each of your customers which not only presents the right product offer but also is timed correctly for when your customer is ready to make a purchase. Also take into account current inventory levels so that you aren't promoting a product that is unavailable or in short supply. Companies like Peak.ai have built enterprise AI platforms that can connect systems that are separated in most organizations and create a powerful connected intelligence system between marketing activities, purchasing processes, and inventory management. Companies using such a system can improve the personalized recommendations made to their customers, forecasting demand, and making sure product inventory is getting to the right markets at the right time.

Erik Larson 's Cloverpop is another example of a company providing such systems (he lists several others in this explanatory piece for Forbes on decision intelligence). Example applications include managing innovation programs, selecting suppliers, and doing better market research. Pratt's company Quantellia provides examples related to climate change, public health, and the justice system.

Where we are: Artificial intelligence is already being deployed by leading organizations today to support better decision making. The two key elements are machine learning models to process the large quantities of data now available and visualizations systems that allow people to make sense of the output from these models. Integrating multiple different information systems can create positive feedback loops -- marketing insights can support better demand forecasting and inventory management for example. Another important dimension is how these systems can bring information to many users at once promoting collaboration and better team decision making.