CRT Unveils Vision-Language Solution 'Vizir.VLM'
A new system combining computer vision and natural language processing to analyze video, understand context, and provide detailed natural-language responses, saving operators over 90% of their time.
«Vizir.VLM»: How CRT Turns Video Analytics into a Conversation with the City
Introduction
On April 21, 2026, at the Securika Moscow exhibition, the CRT Group of Companies (Center for Speech Technologies) presented «Vizir.VLM» — a solution marking a qualitative leap in the evolution of video monitoring systems. Where traditional video analytics merely log events, the new Vision Language Model can "understand" what it sees, establish semantic links, and provide detailed responses in natural language, much like a human would.
This is not just another update to CRT's flagship product. It's the company's answer to a fundamental market demand: video analytics has ceased to be merely a security tool and is becoming the "brain" of urban management. And «Vizir.VLM» is the first step toward making millions of cameras across the country speak to operators in plain language, saving more than 90% of their working time.
Event Details and Timeline
The premiere of «Vizir.VLM» took place at Securika Moscow 2026, the largest industry event in the security sector, held in Moscow from April 21–24. CRT chose this venue to showcase a solution that essentially represents the evolution of their legendary computer vision system, "Vizir."
What's Behind the Name «Vizir.VLM»?
Architecturally, the solution merges two key technologies within a single framework:
- The biometric computer vision system «Vizir» — a mature product which, according to company data, holds over 65% of the Russian video analytics market.
- A Large Language Model (LLM) enabling natural language processing and a conversational interface.
Thus, «Vizir.VLM» receives a video stream or archive recordings as input. Its output, however, goes beyond simple object detection — it can answer complex queries such as: "Find the moment when the driver of bus No. 12 was speeding before a pedestrian crossing" or "Show the waste accumulation statistics in the courtyards of the Leninsky District over the past week".
CRT CEO Dmitry Dyrmovsky described this release as a response to a "new market demand" — the need for proactive management, rather than merely reactive violation logging.
Impact and Significance
The significance of «Vizir.VLM» extends far beyond a single product. It signals a paradigm shift in how cities and large enterprises work with visual data.
For the market and the industry, this means the emergence of a practical tool to realize the concept of "Smart City 2.0." Whereas before, the "Safe City" system (operating on the "Vizir" platform in 30 RF cities) served primarily for surveillance and investigative search, it is now transforming into an analytical decision-making hub. In the housing and utilities sector and urban improvement — which CRT identifies as the most relevant scenarios — the model can not only detect an incident (like a pothole) but also propose a solution (dispatching a repair crew). This shifts urban management from a "firefighting" mode to a state of "proactive management."
For society, the primary beneficiary is the quality of urban life. The claimed boost in city management efficiency by "more than 30%" and a reduction in response times are not abstract figures. Behind them lies real-world logistics: faster snow removal, promptly restored traffic lights, and located missing persons. However, a potential zone of tension also lurks here: the ethical questions of total surveillance, inevitably heightened by any AI-enhanced video analytics system.
For CRT itself, «Vizir.VLM» serves to defend its leadership amid growing competition. The base "Vizir" system is already installed at over 400 transportation hubs and 28 stadiums, including critical infrastructure. The VLM superstructure release creates a significant barrier to entry for competitors: CRT's clients do not need to replace their hardware; they only need to update the "smart" overlay.
Key Player Reactions
CRT's official stance, disseminated through all leading business media outlets (CNews, ComNews, TAdviser), is remarkably clear and ambitious. Company head Dmitry Dyrmovsky acted as the primary spokesperson for the innovation, emphasizing resource savings: "Saving operators over 90% of their time."
The reaction of the professional community at Securika Moscow, judging by indirect indicators (publications in industry outlets "RUBEZH" and "Kuban News"), was positive with a tinge of "long-awaited." The video analytics market has long been "ripe" to stop generating tons of reports that people still end up reviewing.
Analysts and competing developers likely perceived this event as an important benchmark. CRT's starting position (65% market share, presence at critical infrastructure facilities) gives them a considerable head start over startups attempting to implement VLM technologies "from scratch."
Forecast and Conclusions
The presentation of «Vizir.VLM» is not the finish line but the starting point of a new phase.
Short-term forecast (2026–2027): A gradual rollout of the model is expected among key government clients (primarily in "Safe City" projects and on transport infrastructure). The company will gather "field" data for supplementary model training to improve answer accuracy for specific regional queries.
Long-term forecast (2028+): We will see the emergence of industry-specific "Small Language Models" within "Vizir," specialized for narrow tasks: transport security, retail analytics, construction oversight. Integration with "smart contract" systems is possible, where the AI will not only propose a solution but also automatically file a request.
Main conclusion: «Vizir.VLM» illustrates a global trend — the death of "mute video." The era when cameras simply recorded footage so that "if anything happens, we'll check the tape" is fading into the past. It is being replaced by the era of conversational systems, where video becomes a database you can communicate with. CRT has succeeded in being the first in Russia to offer an industrial solution of this caliber, based on its market monopoly. The product's success will depend not on the technology (which is already here), but on the readiness of municipal bureaucratic systems to entrust decision-making to algorithms. Should this happen, «Vizir.VLM» will transform from a surveillance tool into the actual "brain" of urban infrastructure.
— Editorial Team
No comments yet.