As advanced analytics, machine learning, and artificial intelligence rise to the top of many organizations’ agendas, it’s prudent to consider the right balance between technical sophistication and interpretable business value.
Though there is much buzz around analytic models and complex algorithms, the code fueling these aspects is often a small part of a wider ecosystem that needs to work together seamlessly to deliver trusted and actionable insights consistently. Implementations of robust, enterprise-wide analytic models require much different data engineering, architecture, and governance considerations than one-off analytic exercises. Getting these aspects correct is critical to building trust amongst business users. A model that is excessively complex and produces answers that are difficult for the business to interpret detracts from the value of the solution. Appropriately matching the complexity of the model to the business’s needs is paramount.
When embarking on advanced analytics initiatives it’s paramount that the solution is designed to strike the right balance of interpretability, repeatability, complexity, and value. In this recorded webcast, Amanda Teschko explores approaches for ensuring that analytic initiatives are moving users along the curve of sophistication in a way that is sustainable, practical, and trustworthy.