Averian is pleased to announce that it is receiving advisory services and funding from the National Research Council of Canada Industrial Research Assistance Program (NRC IRAP) to support a research and development project to develop an AI-driven assistant that enhances how network and data center operators configure, maintain, and troubleshoot network infrastructure.

The project will focus on the development of an intelligent AI agent leveraging a fine-tuned AI model and document embeddings to assist network operators in performing complex operational tasks with greater accuracy, consistency, and efficiency. The AI agent is designed to operate not as a replacement, but as an assistant to human operators, supporting informed decision-making while maintaining operator control and oversight.

The AI agent will interface directly with a Network Management System (NMS) through mapped APIs, enabling real-time access to network data, diagnostics, and operational controls. By combining AI-driven guidance with live network intelligence, the solution aims to reduce setup time, minimize errors, and improve overall network reliability.

To enhance accessibility and usability for frontline personnel, Averian will eventually develop a user interface optimized for wearable devices such as AI-enabled smart glasses. These hands-free interfaces will allow technicians to interact naturally with the AI agent while working on equipment in the field. 

Security and safety considerations are central to the project, given the solution’s ability to act on behalf of users when executing network operations. The solution will incorporate safeguards, confirmation steps, and operator validation to ensure actions are intentional, traceable, and compliant with operational policies.

Today’s network configuration, operation, training, and monitoring processes still rely heavily on manual input and individual judgment. This can lead to increased costs, inconsistent outcomes, extended setup and repair times, and unplanned network downtime. Variability between operators further compounds these challenges.

By introducing AI-driven guidance, control, and testing, the project seeks to improve consistency, reduce human error, and provide real-time assessments that enable operators and technicians to respond quickly and confidently to network issues. The ultimate objective is to significantly reduce the time required to configure and maintain networks, lower operational costs, and improve the efficiency and reliability of network operations.

Real world use cases include:

• Degraded Performance Troubleshooting
  In the event of degraded network performance, an operator can interact with the AI agent to identify the root cause. The solution queries the NMS, analyzes network conditions, and pinpoints the source of the issue, such as an overloaded switch or allows the operator to instruct the solution to take corrective action, such as reallocating traffic. 
• Commissioning of New Equipment
  When commissioning new network equipment, the operator installs and powers on the device, then scans its label using a wearable device. The AI agent retrieves relevant configuration data, validates connectivity, applies a baseline configuration, and runs diagnostics and operational tests. If issues are detected, the agent recommends corrective steps. Once validated, the agent confirms the equipment is fully operational and generates a detailed commissioning report for future reference.

For more information on Averian’s AI solutions, visit: averian.io