Alteryx Tools Aims to Speed Model Deployment

Data analytics vendors continue to push deeper into organizations with tools designed to address the pressing problem of actually getting predictive models deployed into business operations.

Among those promoting the rise of the “citizen data scientist” is self-service analytics specialist Alteryx Inc., which this week unveiled a new component to its flagship analytics platform enabling users to deploy predictive models and manage performance over time via an API.

The company (NYSE: AYX), said Tuesday (Sept. 12) its new tool dubbed Promote stems from its June 2017 acquisition of Yhat, developer of a machine-learning platform for accelerating deployment of “real-time decision” APIs.

The API approach is touted as helping data scientists overcome the “last mile” problem when it comes to transitioning predictive models to production. The company added that its API approach eliminates the need for IT support when deploying and updating predictive models.

The tool works by embedding machine learning and predictive models into production applications. Those apps are capable of accelerating use of REST APIs without recoding. Models can then run on-premise or in the cloud.

The Alteryx approach is designed to address a growing problem for data scientists, namely, the inability to get their predictive models deployed in production. The company cited survey results indicating the only 13 percent of data scientists succeed in deploying their models, a total that hasn’t improved much over the last eight years.

Alteryx claims its tool can deploy models in minutes, and the resulting API requires only a single line of code that can be incorporated into HTML web sites. The result, the company claims, is predictive analytics without coding.

“The next level of maturity is operationalizing those models for consumption,” Ashley Kramer, vice president of product management at Alteryx,” explained in a statement. “This convergence of machine learning, business intelligence and data science is essential for modern organizations to realize the full potential of their data assets.”

Alteryx CEO Dean Stoecker asserted that the deal for Yhat would help make modeling tools more accessible to speed deployment of advanced models. The self-service tools are aimed at data scientists who often must rely on open-source programming languages to build predictive and machine learning models. The challenge is that open source statistical tools are often incompatible with frameworks and languages used to build applications.

Those models must then be deployed into different technology frameworks, a step Alteryx cites as a bottleneck in deploying models.

“The Yhat platform helps reduce the roadblock between data scientists and development teams by accelerating the model making and model deployment processes,” the company said.

Yhat had touted its approach as requiring fewer development resources to get analytic models out the door faster.

Alteryx Promote is scheduled for general availability in early 2018 as an add-on to Alteryx Server.

Recent items:

Alteryx Buys Yhat, Rolls ‘Connect’ Platform

Alteryx Goes Public as Data Prep Market Heats Up