Earnix, the global provider of intelligent, composable, SaaS solutions for insurers and banks, announced Model Accelerator at Earnix Excelerate 2023. Model Accelerator is a web-based module designed to streamline and accelerate the process of building and incorporating advanced models in pricing, underwriting and real-time rating.
One hundred percent of insurance leaders plan to use machine learning models in their pricing and underwriting processes, according to a 2023 study commissioned by Earnix1. However, only 20 percent of the survey’s 400 respondents said they do so today. Factors leading to slow adoption include difficulty importing external models; the iterative and manual processes required to build generalized linear models (GLMs); and lack of flexibility in machine learning model development due to software constraints. Additionally, it takes time to see the value of new machine learning models because the complex models need to be recoded into the rating engine and extensively tested.
With these needs in mind, Model Accelerator was created as the first tool to offer model import, build, and assist capabilities in one platform. A new addition to Earnix Price-It and Underwrite-It, Model Accelerator offers integration accelerators, allowing integration with the most popular frameworks and formats with ease; modern approaches to kickstart model building, such as Earnix Automatic GLM (AGLM), and hybrid modelling.
Unlike other rating engines, Earnix enables insurers to bring in any model they develop and use it in production across pricing, underwriting, and rating. These elements significantly reduce the complexity around interoperability of models and open the opportunity to create modern, cutting-edge, externally developed models and still use them, natively, within Earnix.
“Operationalizing machine learning models is complicated and complex, and our customers want the flexibility to choose the best analytical approach to fit their needs,” said Yaron Lavie, Vice President, Products, Earnix. “By expanding open analytics capabilities, streamlining processes to use machine learning models in production, and providing new techniques to build models faster, Model Accelerator empowers our customers to take sophisticated modelling to the next level.”
Model Accelerator increases transparency and explainability, giving business, underwriting, and pricing leaders and regulators a clear understanding of how models are used. It enables pricing, underwriting and business professionals to understand the key variables driving predictions and how changes to those can affect predictions of the model, and effectively leverage these models for making more informed business decisions.
