Frontline Systems Releases RASON V2023 Cloud Platform with Advanced Model Management and Decision Intelligence

Frontline Systems is shipping RASON V2023, a new version of its cloud platform for advanced analytics, that enables both business analysts and developers to easily create and run models using mathematical optimization, Monte Carlo simulation and risk analysis, data mining and machine learning, and business rules and calculations.

RASON is not new – since 2015 it has been available “around the clock” as an Azure-based SaaS platform, one of the first to offer a REST API for both predictive and prescriptive analytics. RASON has been enriched each year with new analytics and decision intelligence features – including support for business rules using the DMN open standard in 2019. Now, RASON V2023 features new tools to manage and govern predictive, prescriptive, and decision models, assess the risk of deploying trained and validated machine learning (ML) models, and track runs of all types of models in production use.

“With a patent application now on file to preserve invention rights in our machine learning risk analysis methods, we’re releasing an array of new capabilities for analysis, management and governance of analytic models”, said Daniel Fylstra, Frontline’s President and CEO. Frontline Systems is concurrently releasing new versions of Solver SDK®, its object library for developers, and Analytic Solver, its tool for business analysts using Excel for the Web, Windows and Macintosh, with support for the same risk analysis methods.

Recommended AI News: Complete Verkada Platform is Now Available Across the UK and Europe

High-Level Analytics Modeling Language Embedded in JSON

RASON is used to create and solve virtually any type and size of optimization model, up to millions of inter-related decisions; perform risk analysis for virtually any type and size of Monte Carlo simulation model; train, validate, and test a wide range of forecasting and machine learning models up to large scale; create and execute business rules and decision tables that use analytic model inputs; and perform general calculations. RASON makes it easy to execute models and retrieve results “on demand” via a REST API, with results in easily-consumed JSON and OData form.

RASON provides an all-inclusive, high-level language for analytic models, embedded in JSON – making it easy to use in JavaScript, Web and mobile applications. Within RASON (“inside quotes”), users can directly use four familiar open standards: business rules in DMN (Decision Model and Notation) and FEEL (Friendly Enough Expression Language), machine learning models in PMML (Predictive Modeling Markup Language), probability distribution models in SIPMath 3.0 from ProbabilityManagement.org, and the Microsoft Excel expression language, using virtually any Excel formula or built-in function. This enables organizations using RASON to leverage the knowledge and experience of business analysts, data scientists, and Web developers, and enable these diverse users to work together – avoiding the “re-invention of analytic models” so common in today’s world.

RASON makes it easy to connect models to data, ranging from CSV files and Excel workbooks to SQL Server on Azure, OneDrive for Business, the Common Data Service (CDS) underlying Dynamics 365, Power Apps and Power Automate, and CData’s Cloud Hub with access to over 100 other cloud data sources – using Azure Vaults for access credentials.

Synthetic Data Generation and Risk Analysis of Machine Learning Models

Synthetic Data Generation (SDG) has become topical in machine learning in recent years, with a number of companies founded just to supply software and services around this technology. SDG is used when there isn’t enough original data, or when use of the original data is restricted by law or regulation. But until now, SDG has simply been used to better train ML models.

RASON V2023 includes a powerful, general-purpose, easy to use Synthetic Data Generation facility. Unlike some special-purpose SDG offerings, this facility can accurately model the behavior of nearly any combination of features with continuous values. But RASON also uses synthetic data in an entirely new way, to analyze the risk that a ML model will yield unexpected results “large enough to matter” when deployed for production use.

Recommended AI News: Blotout Announces New Partnership with Fastly to Improve Meta Ad Spend with Blotout’s EdgeTag

Risk Analysis Works with Already-Available ‘Augmented Machine Learning’

RASON’s V2022 release featured “augmented machine learning” features found only in other sophisticated machine learning tools. The user simply supplies data; RASON will test and evaluate multiple types of machine learning models – classification and regression trees, neural networks, linear and logistic regression, discriminant analysis, naïve Bayes, k-nearest neighbors and more – validate and compare them according to user-chosen criteria, and deliver the model that best fits the data. With a command as simple as “simulation”: { “sampleSize”: 5000 }, the user can perform a risk analysis on the “best model” found, to quantitatively assess how it may perform differently on live data in production.

From Single Models to Deployment, ‘ModelOps’ and Decision Flows

Users seeking to move to the next stage of ‘MLOps’ or ‘ModelOps’ find that RASON makes it easy to deploy models, with automatically-created REST APIs. But RASON goes much further, making it easy to define multi-stage “decision flows” that can include data retrieval and in-memory SQL queries; training, validation, risk analysis, and use of machine learning models; business rules and optimization models; and delivery of results. RASON results are available in JSON and OData formats, for use in popular BI and analytics tools like Power BI, and low code / no-code application development tools like Power Apps and Power Automate. This sophisticated combination makes it easier than ever to build powerful, automated decision intelligence solutions.

Model Management and Governance with Organization Accounts

RASON supports multiple models, with built-in versioning and use of “champion” and “challenger” models. Organizations can use a “RASON Organization Account” – enhanced in V2023 – to maintain all of their users, models and data in their own secure, private Azure Storage account, using Azure’s User Authentication, Azure Vaults for access credentials, and Role-Based Access Control (RBAC) for containers and models; RASON itself runs with only the permissions an admin grants to a given user.

New in V2023, RASON includes rich facilities for cataloging, managing and governing models of different types (machine learning, optimization, simulation) and formats (RASON, PMML, etc.), models created by different authors or those set for “development” or “production” use. RASON makes it easy to view summaries and detail of runs of those models including time to run, CPU resources used, errors and more. Every RASON user, but especially larger teams, can use these facilities to evaluate their progress and demonstrate how their work is being used to realize business value.

Recommended AI News: Raising the Ceiling of Code-Free Programming for Robotics and Industrial Automation

[To share your insights with us, please write to sghosh@martechseries.com]