Built For Continuous Improvement

In the business world, people take action with the specific goal of improving performance. But taking an action such as changing a price, has additional value as an experiment that provides new data about the market. In order to learn from the experiment, you have to measure the results and compare them to your prediction.

This is easy to do in a lab, but tough in the real-world where changes in revenue and profit are affected by more than just the price. Predictive models must have the automated intelligence to learn from actual results or they will quickly fall short. Markets are too volatile and complex for a static or manually maintained system to keep up.

Zilliant science has the built-in capability for closed-loop continuous improvement. In science terms, these capabilities are called machine learning. Learning goes on across several aspects of the system, especially in the model parameters, where our algorithm filters out true feedback signals to make ongoing refinements.

Of course for Active Learning to occur, the model has to have ongoing access to updated transaction results. Zilliant adopted the Software-as-a-Service model, in part, to keep the loop closed between the science and ongoing results.