Shield Wires OpenAI's GPT-6 Astra Into Forge to Let AI Agents Navigate Legacy IT Portals

Shield, a company that makes software for IT services providers, says it has connected OpenAI's GPT-6 Astra to its Forge platform. The result, according to the company, is that AI agents can now log into aging customer portals and finish the routine browser work that has long tied up human technicians.

Shield says the workflows are not a pilot. They are running in production, with thousands of executions each month across its network of managed service providers. Forge is pitched as an AI operating system for IT support shops. It pulls together business data and customer context so that a ticket can be carried from start to finish.

The problem the integration targets is access. Many of the portals technicians use every day to manage tickets, devices and software licenses predate modern APIs. An agent working in them has to read the screen and click through it the way a person would, an approach known as computer use. Shield's earlier attempts at it were not dependable enough, according to Mayank Jain, the company's Head of Technology. "Usually, a system that doesn't have an API is an old-school system that also doesn't look very nice and is very hard to use, even by a human," he said.

Under the new setup, Forge's AI Engineer hands supported browser tasks to a computer-use agent. Technicians set the limits in advance: what the agent may do unsupervised, and what must come back for review or approval. Shield says the agents now complete those steps themselves instead of leaving a customer waiting for a person to log in. "The only time you should have to bring in a human technician is if you need their judgment, not because you're blocked on access," Jain said.

Shield reported several performance results. Tasks finish 35-40% faster on average than with earlier models. The company also cites fewer failures and fewer timeouts, less time spent by technicians writing detailed agent instructions or guiding tasks, and higher customer satisfaction on tickets where AI is used. These are Shield's own measurements of its own platform, and they have not been independently verified.

Jain described the handling of interfaces as the biggest departure from earlier generations, which visibly stumbled while moving through a screen. "GPT-6 Astra feels like watching a human click around," he said.

The division of labor is not total. Phone calls, on-site visits and remote-desktop work still require people. Shield is testing remote-desktop workflows that would let agents take on more complex actions directly on a machine's desktop.

The timing matters for small and medium-sized businesses, which typically rely on outsourced providers for everything from a locked account to a misconfigured firewall that has knocked an office offline. Speed of response is what those providers sell, and any step that stalls on a portal login shows up as a delay. Shield argues that handing routine navigation to agents frees technicians to concentrate on higher-priority incidents such as phishing attacks or critical outages. Jain said the agents can now perform tasks that only a human could handle six months ago, and that Forge is now "truly an extension of our human technicians."

For IT services firms, the competitive question is shifting from whether agents can act to how much of that action can be trusted without a human watching. Shield's answer is the permissions layer. What to watch is whether technicians widen the set of tasks they let run unattended, and whether the remote-desktop tests move into production.