VK Predict, part of the VK group, has launched AutoML, a platform for automated machine learning model development. This tool aims to assist business analysts in devising data-driven strategies and tackling marketing challenges, while also enabling collaborative AI model training without data sharing through federated learning. This feature is particularly attractive to clients who cannot exchange sensitive data due to commercial confidentiality.
AutoML provides business analysts with capabilities to assess market positions, segment audiences, identify cost-effective clients, and personalize communications. The platform supports various tasks, such as forecasting performance, ranking entities, and clustering data. Users can input datasets that are processed and analyzed by the AutoML platform, which then selects and trains suitable machine learning (ML) models to produce actionable insights.
A noteworthy aspect of AutoML is its ability to facilitate federated learning, allowing organizations to train models collaboratively without exposing their raw data. This is particularly beneficial in sectors like e-commerce and fintech where data confidentiality is crucial.
VK anticipates that the demand for low-code and no-code platforms will grow, with a significant portion of businesses projected to adopt these tools for application development by 2026. The target audience for AutoML includes companies of all sizes requiring data analysis and forecasting capabilities.
While AutoML solutions can enhance and automate certain aspects of data science, experts emphasize that they should not wholly replace data scientists. Instead, they serve as a valuable complement, expediting the application of ML methods. The platform’s effectiveness will be gauged by its ability to handle various tasks efficiently and appeal to non-specialists in IT.