NNTC integrates intelligent video analytics solutions that detect and recognize separate objects (people, vehicles, zones, etc.) in video, monitor behavior, and make forecasts. The proprietary runtime implementation of neural networks enables face processing to be executed in a fraction of a second, outperforming popular platforms in blind tests. Modules work independently from each other. In combination with other technologies such as virtualization, NNTC builds powerful, distributed and scalable solutions.


Video analytics solutions help retailers run efficient marketing campaigns,   improve efficiency of shopping zones, and deliver better customer experience.

Video analytics allows you to:

  • Count unique visitors and identify your loyal customers once they enter the store

  • Assess the efficiency of your marketing campaigns through online sociodemographic analysis and visitor counting

  • Analyze shopping zones to use your space more effectively

  • Monitor queues to react on peak times promptly and optimize queue time

  • Deliver personalized advertising right at the point of sale

Use case: Retail


  • Employee on-work time registration

  • Detection of suspicious persons entering premises (black listing with security alerts)

  • Control over POS terminals

  • Reports and forecasts for peak times, number of people entering/leaving the store, and so on

  • System is rolled out in two stores now

  • By the end of 2019, it will be rolled out in 80 stores across the country


NNTC’s proprietary face recognition engine detects, captures, and  recognizes multiple faces per second and can be effectively used by law enforcement   officers in crowded areas and at critical infrastructure  facilities (transport hubs, energy/water/heating  supply facilities), thus ensuring proper security on sites where individuals are to be identified or registered, including:

  • Entrance areas of transportation stations, airports, subways,  sport centers etc.

  • Border customs control

  • Video surveillance, verification and access control

The system automatically selects the best position for face capturing, transmits it and finds a match in the database, regardless of changes  in skin, hair, etc. The system significantly reduces operator's load: no need for constant monitoring of all control areas and no space for human error.

Use case: Public Security


  • AI face recognition surveillance system covering the whole region

  • 21 cameras installed in the Sakhalin Region checkpoints to recognize wanted individuals

  • 98% of people identified – the highest possible efficiency on the market

  • All departments have access to the system

  • The system deployed on 12 sites (airports, ports, railway stations, etc.)

  • Anti-terrorism control

Read more:


Time and attendance

Analytics of video footage allows company’s management to gain valuable insight into employee attendance and generate various custom reports with a large set of filters. As cameras capture employee entering the workspace, a facial recognition system instantly identifies an individual and changes their attendance status.

Intelligent visitor management

Additionally, face recognition can be a basis for intelligent visitor management solution — a software and hardware system for logging visitors.

  • Convenience and economy of time: a visitor pass can be requested and approved from the workplace, as tablet or smart phone allows the employees to use their work time more efficiently

  • No queues at the reception: pass issue in less than a minute

  • Face recognition: guest movement tracking at all entrances

  • Kiosk: optional self-registration kiosks to reduce traffic during peak loads

  • High security: pass approval by different responsible persons, blacklists and site visit reports

  • Simple integration and servicing: web client use, integration with an enterprise portal and directory services, and Windows authentication


    NNTC is a supplier of AI-assisted video monitoring products by iCetana, which plugs into VMS, DVR, or cameras directly. The system analyzes every camera feed 24/7, learns what’s normal in a scene and identifies anything abnormal. Once an abnormal event happens, it is displayed on a monitor, attracting operator’s attention if necessary and fostering a faster reaction.


    • Undesirable and abnormal events detected first
    • Self-learning system adapting to change
    • Performance in real-time and at scale

    How it works

    • Camera feed analysis: objects, edges, contrast, motion, direction, and speed
    • Automatic identification of standard patterns
    • Anomaly detection before the situation gets worse
    • Abnormal event displayed and incident details instantly available for better response and reporting
    • Operators taking action when necessary
    • Real-time and historical reporting and data visualization