Smart Monitor Adds Tools for Managing AI Infrastructure
The AI Observability module gives companies a single-pane view for monitoring intelligent systems.
Russian companies have increasingly been deploying AI infrastructure on their own systems. That is creating demand for software that can help companies operate intelligent modules securely. The latest update to VolgaBlob’s Smart Monitor platform focuses directly on that need.
The Update Fills a Key Gap
According to the developers, conventional infrastructure monitoring does not account for the specifics of AI: it does not track GPU memory usage, cost per request, AI agent traces or AI service availability. The latest version of Smart Monitor is designed to fill that gap.
Smart Monitor 6.1 introduces a component called AI Observability. It collects telemetry from multiple sources across a company’s AI environment and consolidates it into a unified, out-of-the-box monitoring layer. The system tracks the status of GPUs, proxy servers and inference servers. It also monitors performance, resource availability and token consumption.
Together, these capabilities give companies a comprehensive view of how their AI infrastructure is operating, including its performance and costs. The single-pane approach can also make it easier to identify failures and address them before they cause financial or operational damage.
Keeping AI Agents Under Watch
The AI Security module is another notable part of the platform. Its primary role is to detect threats associated with corporate AI tools. The module monitors the actions of local AI agents as well as prompts sent to and responses generated by large language models. The updated version adds new security rules aligned with data collection configurations and developed in accordance with the OWASP Top 10 for AI standard.
Another significant addition is ML-studiya (ML Studio), a framework designed for machine learning tasks and now built directly into the platform. ML-studiya lets users work with a ready-made forecasting model, allowing engineers and analysts to handle ML tasks without leaving Smart Monitor. It also includes a library of default algorithms, with the option to add custom ones.
The Update Goes Beyond AI
The new AI capabilities are only part of the release. Smart Monitor’s core has also received substantial changes. The base Core module now includes a streaming correlator that processes data as it arrives. The Mayak (Beacon) module adds a maintenance mode that suppresses false alerts from systems undergoing technical work. Meanwhile, the task scheduler can now distribute launches over time: it analyzes the schedules of all active jobs and suggests optimizations. This prevents hundreds of jobs from starting simultaneously and creating peak loads.
A New Market Segment Is Taking Shape
Smart Monitor’s update history across 2025 and 2026 shows the platform gradually evolving into a general-purpose information security tool spanning multiple areas.
AI management tools, meanwhile, are beginning to form a distinct market segment where developers compete for customers. In March 2026, Yandex B2B Tech launched its own observability platform, Yandex Monium. It can monitor the health of applications, digital products and AI agents. The system is designed to handle massive workloads by running its infrastructure on fault-tolerant hardware deployed across Yandex data centers.
More tools for securing AI components are likely to enter the market soon as their use within corporate environments continues to grow. Rather than stand-alone services, these products are likely to take the form of general-purpose platforms capable of handling a broad range of tasks.
