Compellon – Assumption Free AI for the Analyst

Our proprietary IPC 20|20 AI platform deviates from error - prone conventional machine learning approaches that use an assumption - based trial - and - error methodology to best simulate the behavior in enterprise data. Instead, we use an assumption - free approach that starts with the business analyst’s definition of the problem and then automatically invents a model structure that directly captures the data’s behavior to that problem. More importantly, the Compellon platform is adaptive – it continues to learn and enhance its solution post - deployment, seeking ever - improving solutions in each operationalized iteration. Our solution has been successfully applied to business problems across multiple verticals, including healthcare, insurance, manufacturing, CPG, media & telecom, banking, retail, government and others. Representative use cases include merchandising optimization for eCommerce, reduction of mean time to problem resolution in data center operations, reduction of readmission rates in health care, subscriber churn reduction and upsell strategies for media services, and increasing effectiveness of collection tactics in the financial sector. IPC 20|20 is ideal for when you need to know if your data supports your business objective. Using measures of information and consistency, IPC 20|20 evaluates your data’s ability to build a stable predictive model for your target. Based on these assessments, and your business objective, you can then decide to proceed, get more data, or change the scope of the project. To use IPC 20|20 , all you need is: A defined business problem or question – “How can I reduce our churn rate?” Data in a rectangular format with the target variable* included – Churn And a Subject Matter Expert – that’s you!