Network operations (NetOps) solutions provider NetBrain Technologies has expanded its agentic automation platform to help engineers diagnose and remediate faster.

Following a series of updates in June, NetBrain has added the ability to build custom AI agents. Available from early October to all customers, engineers can design and deploy autonomous agents for specific tasks like proactive fault identification.

Customers can also make use of preset agents, with NetBrain adding agent offerings it claims can perform change and assessment tasks to evaluate infrastructure security and performance. The platform already offers preset agents that remediate network issues and validate network paths.

“Network engineers already use the live context and automation on our platform to reduce change execution and validation effort by 85%," NetBrain CEO Bernadette Nixon said. “Now, with a growing team of agents that can take on any NetOps task under human governance, they can extend that same impact across the full lifecycle, from diagnosis to decision to action.”

Agentic AI has quickly become the hype technology of network automation, with vendors touting it as a means to automate complex and time-consuming tasks. NetBrain has taken the concept of agentic AI and run with it, making it a central point of its offerings alongside its more traditional automation means.

Despite doubling down on agentic offerings, the vendor was keen to position its agents as working with engineers in the loop. NetBrain said agents and humans can “reason and collaborate on live network context, cutting the time between an alert and a confirmed fix.” NetOps Harness acts as the vendor’s guardrails, paired with agents to ensure they only use topologies, paths, and device data already trusted by their human counterparts.

“Customers want agentic reasoning that's grounded in the live state of their network, not an outdated model, combined with safe, human-approved automation," said Song Pang, NetBrain’s CTO. “That requires a best-of-breed NetOps Platform and Harness that enables and manages agents optimized for network operations. And with the model context protocol, a customer can bring any AI tool of their choice into their environment and enable it with our live network context and automation.”