NVI Solutions participated in the Data Fusion 2025 conference, dedicated to big data and artificial intelligence. The event took place on April 16–17 in Moscow at the Lomonosov Technology Cluster and brought together leading experts from the IT industry, business, academia, and the public sector.

Data Fusion is a cross-industry platform that fosters dialogue between business, science, and government, focusing on current issues related to data management and the development of AI technologies. The conference featured over 70 sessions covering topics such as the platform economy, personal data protection, government regulation of digital services, and case studies and research in AI and neurotechnologies.

As part of the conference, an NVI Solutions representative delivered a presentation titled "Personal Protective Equipment Recognition in Industry: Development, Scaling, and Adaptation for Harsh Arctic Conditions and Low Visibility" during a session dedicated to computer vision.

The speaker outlined two key approaches to building machine learning systems. The first focuses on efficient integration between trainable neural networks and non-trainable algorithms. The examples demonstrated how high-quality results can be achieved when a neural network is tasked with solving a specific problem, while the final decision is made by an interpretable algorithm that corrects potential errors made by the AI. The second approach involves creating synthetic datasets. The case study showed how high-quality training data for neural networks can be generated with minimal use of real-world data and without significant costs for manual labeling.

Data Fusion 2025 served as a unique platform for business, science, and government stakeholders to exchange ideas and discuss the future of artificial intelligence. NVI Solutions’ participation in this major event highlights the company’s commitment to innovation and technological advancement in the fields of big data and AI.

‍