Not long ago, it was hard enough to build models that applied data science and machine learning, let alone advanced analytics like simulation and optimization. But it was so difficult to move from a working analytic model in a custom tool, to a model that was deployed for many users, usable by other applications, updated with new data, monitored and managed, that companies just never “got this done”.
But today, this can be easy, if you’re using Microsoft’s Office 365 and Power Platform – and even better if you’re using Dynamics 365. From single planning models to an entire “analytics center of excellence”, RASON can simply your job by leveraging what you already have: business expertise and Excel skills for model creation, low-code/no-code tools such as Power Apps and Power Automate for model deployment, and Common Data Service, OneDrive, OData and CData tools for data management.
RASON – an acronym for RESTful Analytic Solver Object Notation – is our Azure-based platform for deploying, using and monitoring analytic models the easy way. Consider:
- Every RASON model has a REST API, with JSON requests and responses – easy to consume in any web or mobile app, from C# to JavaScript.
- Every RASON model has an OData endpoint for results. Tools like Power BI and Power Apps can consume these results directly – no code required.
- Every RASON model can use a wide range of “pluggable” data sources, from Excel and CSV files to SQL Server, the Common Data Service, or 100+ CData Cloud Hub sources.
- RASON makes it easy to build multi-stage decision flows of SQL operations, data transformations, machine learning, optimization, simulation and business rules.
- Every RASON model has tracking of versions as well as runs – and since RASON monitoring has its own OData endpoint, you can view, slice and dice the history in Power BI.
It’s so easy to create and deploy analytic models using RASON and Analytic Solver that line-of-business analysts can do it themselves, without needing, or placing a burden on central IT. And when that initial project blossoms into a tool that should be extended and managed by IT, developers can work easily with RASON’s clean, fully open, documented REST and OData APIs.
Even better, “hand-offs” from business users to developers need not be “one way, one time”: Thanks to RASON’s model interfaces and its support for both DMN FEEL and the Excel formula language, business analysts and developers can work side-by-side on an analytics project, with RASON as their “lingua franca”.
And with our forthcoming V2020.5 release, this whole deployment story is about to get even better. We’re making everything easier, from “how to move from Excel to RASON”, to “how to build multi-stage decision flows”. Most of all, we're making it easier to use and re-use every analytic model you create, so you get the most value from your analytics modeling efforts. Stay tuned!