- Russian developer of AI systems for video and speech analytics and AI agents. Products are listed in the Russian software register.
- 40+ industrial deployments, 25,000+ connected recognition channels; 80% of key customers come back for new solutions.
- Monitoring of 1.2M+ waste container sites across Russia for the Russian Environmental Operator; near-real-time transcription of 12,000+ audio channels for Fix Price.
- Up Great technology contests: 1st of 180 teams in Autonomous Search (2024) and 2nd of 240 in Emergency Search (2023) — onboard detection of people in drone imagery.
CVS products
- AIVA
- server-side video analytics for plants, retail, pickup points and restaurants
- VION
- finding violations in photos and video from mobile devices and drones
- NOVA
- on-device analytics for cameras and terminals, food recognition
- Forge AI
- configurable AI assistant: RAG, documents, speech analytics
- FRS
- non-cooperative face recognition integrated with access control
- NPS
- neural processing server for drone imagery
- 100+ staff in 6 directions: machine learning, applied neural networks, distributed systems, AutoML, reinforcement learning, LLMs and AI agents. 25–26 projects a year.
- A domestic neural network training platform for the Ministry of Industry and Trade: Platform-GNS → Automat-GNS → Fundament-GNS. Certified source code, support for Russian hardware, 300+ registered results over 9 years.
- Scientific and technical lead of Advanced Research Foundation projects: Marker (vision for an autonomous robot), Paradigma (AutoML), Taktik (reinforcement learning).
- A virtual co-pilot: pilot gaze and fatigue tracking, a crew voice assistant with 200+ commands at 95% accuracy; onboard predictive diagnostics for aircraft.
- Author of a course for managers: leadership, teams, company life cycle, CJM, P&L.
- Cross-platform biometric core and the Biometric Access Control client-server system; commercialisation of research.
Associate professor
Moscow Aviation Institute
2009–2014, 2021 — present
- Probability and statistics, image processing, artificial intelligence and machine learning, applied statistics.
Expertise
- Video analytics and biometrics
- face recognition at checkpoints and in crowds, anti-spoofing, PPE control, behaviour analytics, retail theft, marketing metrics
- Drone and aerial imagery
- power lines and towers, pipelines, cadastre, landfills, invasive plants, forest inventory, search for people, roads
- Speech, LLMs and AI agents
- speech-to-text analytics, RAG assistants, document generation and review, requirements analysis
- Robotics and autonomy
- neural SLAM, stereo depth, camera and lidar segmentation, traversability maps, ROS2
- AutoML and reinforcement learning
- architecture search, pruning and quantisation, RL for adversarial games and control
- Onboard and trusted AI
- optimisation for CUDA, TensorRT, OpenVINO, ARM and Russian hardware; platforms with certified source code
Technologies
PyTorch, CNNs and transformers, Conformer, BERT, LLMs, RAG, MCP · C++17, Python, C#/.NET, React, Qt, OpenGL · CUDA, TensorRT, OpenVINO, MNN · Docker, Kubernetes, Slurm, Ansible, ROS2 · Agile, Azure DevOps, GitLab, P&L, CustDev, CJM
