In this blog post we describe SMARTS’ key capabilities: Decision Management Platform, Low-Code / No-Code Apps, Machine Learning Ops, and Real-time Decision Analytics.
SMARTS is a decision management platform that spans the entire business decision lifecycle, from modeling to deployment. Decision Management Platform, as the core functionality of SMARTS, includes authoring, testing, simulating and deploying decisions as well as the management and governance of those decisions. Key authoring capabilities include RedPen™>, which allows business analysts to write rules via point and click without learning a special syntax, and Fluid Metaphors, which allows business analysts to choose the most appropriate decision representation based on the task at hand. The same decisions can be viewed and modified in the form of decision rules, tables, trees, or graphs.
SMARTS is a data-centric product in that business analysts can import data and immediately start writing and testing rules against that data, making the building of rating, pricing, and any calculation engine an easy task. SMARTS’ powerful processing of spreadsheets allows business analysts to retrieve rates from a very set of options; fine-tune the rates to cope with exceptions and overrides; and manage these rates over different time horizons, per geography, any user-defined dimensions.
SMARTS lifecycle management capabilities include release management (read-only snapshots of projects) to track promotions and deployments. Business analysts can view the history of their decision logic and revert to earlier versions or releases ant anytime. Based on their permissions, they can publish their changes into a test environment or into production.
SMARTS decisions are typically deployed as decision services either on the cloud on on-premise. SMARTS decision services provide support for secure service invocations in an authenticated context, support for high-availability and scalability.
All of SMARTS other capabilities, Low-Code/No-Code, Model Ops, and Real-time Analytics are integrated into to the Decision Management Platform.
Business analysts can develop decisions quickly and intuitively with SMARTS’ Low-Code capability. SMARTS Low-Code features include graphical representations of decision logic such as decision flows, tables, trees and graphs. Rules are written using the SparkL (pronounced “sparkle”) user-friendly language or with the point-and-click Red Pen feature.
Thanks to this low-code capability, business analysts can develop Business Apps which provide a no-code environment for untrained business users and ‘citizen developers’ to modify, test and deploy changes.
Many decisions deal with uncertainty- such as credit card approvals and loan and mortgage approvals. These decision benefit from predictive models that help organizations make the best possible decision in the face of uncertainties. SMARTS operationalizes predictive models in three ways.
- Business analysts can import machine learning models developed by a data science team using PMML (the standard model interchange format). Once imported these models can be integrated into any decision and deployed. The models can be tested in the context of the decision to ensure they produce expected results and then can be deployed.
- SMARTS’ patented BluePen machine learning technology can be used by business analysts to generate rules from data. Models created by BluePen™ are explicit and easy to understand.
- In cases where models exist but are not available as PMML, but are available as spreadsheets, SMARTS can import them as a dashboard, regression, or tree model.
SMARTS Real-Time Decision Analytics lets business analysts define metrics that measure the outcome of SMARTS decisions to ensure they are supporting business strategy and objectives. The SMARTS dashboard displays graphical representations of the metrics allowing the business analysts to analyze the impact of changes to decision logic and run simulations against large volumes of historical data. SMARTS decisions can also be monitored post-deployment and alerts can be defined which notify stakeholders when thresholds are exceed.
Using SMARTS Real-Time Decision Analytics capabilities, Business analysts can measure the end-to-end performance of the decisions that invoke these models – not only the quality of the model but the quality of the whole process.
We hope describing SMARTS in-terms of its powerful capabilities has given you a better understanding of the platform. Insurance companies, banks, and healthcare companies rely on decisions powered by SMARTS every day for loan origination, claims processing, credit risk assessment, as well as fraud detection and response.
Learn more about Decision Management and Sparkling Logic’s SMARTS™ Data-Powered Decision Manager