Moscow Region officials report significant operational gains from deployed machine learning models across local healthcare infrastructure. Governor Andrey Vorobyov highlighted progress under the "Data Economy" national project, emphasizing automation designed to reduce administrative load rather than replace clinical judgment.
The region's voice assistant, "Svetlana," now handles peak loads equivalent to 600 human operators. In 2025 alone, the system processed over 8 million incoming and 9.1 million outgoing calls, managing appointments and proactive prescription renewals. This NLP deployment frees staff from routine scheduling, recovering thousands of lost appointment slots caused by no-shows. By automating confirmations, the system ensures capacity isn't wasted on absent patients.
On the clinical side, speech-to-text models trained on specialized medical terminology—supporting both Russian and Latin—have cut diagnostic reporting time by half. Victoria Ten, Chief Doctor at Lobnya Hospital, noted interpretation times dropped from 45 minutes to roughly 15 minutes across ultrasound, X-ray, and MRI departments. Results upload directly to patient portals, removing physical retrieval steps and accelerating the data pipeline to the end user.
Decision support tools are also active in production. Prescription assistants now validate medications against 15 risk parameters, checking for interactions and dosage limits during both in-person and telemedicine visits. Additionally, a diagnostic assistant offers secondary opinions based on electronic health record data, though final authority remains with the specialist. Patients can also view AI-interpreted lab results via the "Dobrodel Health" app, receiving plain-language guidance on deviations while awaiting professional review.
The initiative demonstrates how targeted model deployment can optimize workflow efficiency without compromising oversight. Engineers focused on integrating these tools directly into existing electronic health records, ensuring seamless adoption by medical staff who report high satisfaction with the time savings.
Source: RIA Novosti