MWS AI, a division of MTS Web Services, has launched a corporate data labeling platform designed to train artificial intelligence. The solution replaces foreign tools including LabelStudio Enterprise, Prodigy, Supervisely, labelit, and Audino, covering over 99% of industry scenarios for text, audio, image, and video processing.
The platform features a web interface with a low-code builder and an AI agent that allows users to customize projects using natural language. It supports deployment on customer infrastructure, including critical information infrastructure, and is currently undergoing registration in the Russian software registry.
The system enables automated labeling by connecting to large language models via API. It also supports manual labeling with model-assisted suggestions and multi-user verification workflows. Results can be exported for model training and evaluation purposes.
In 2025, the transition to this platform reduced MWS AI labeling costs by 25% and saved over 17,000 man-hours. During the first 10 months of 2026, the company saved 30,000 man-hours. Data preparation speed increased by 30%, and the onboarding time for new labelers dropped by 35%.
The platform is currently used by MWS AI, MTT (Exolve), and the MTS Group. It is available as a standalone product or as part of the MWS AI Agents Platform. Customers can choose to manage labeling with their own teams or outsource the work to MWS AI specialists.
