Orion soft has released an update to its StarGuard AI platform to version 1.1, expanding the tools for controlling corporate use of large language models. The release includes monitoring, expense tracking, and analytics tools, as well as a redesigned user interface.
One of the main innovations is the AI request billing mechanism, which allows tracking the consumption of language models by users, departments, and model types. This helps analyze token usage and distribute costs among teams.
The platform also features analytical dashboards for tracking information security indicators, user activity, and AI agents. In real-time, it is possible to monitor model access and compliance with corporate policies.
In addition, the interface and navigation have been updated, automatic user registration upon first login has been implemented, and a centralized registry of AI agent tools has been added. The new version expands sensitive data detection mechanisms, improves semantic analysis, optimizes the processing of requests with large contexts, and introduces a sliding window mechanism for managing user limits.