Network verification company Forward Inc. today launched Forward Predict, a new capability that lets network teams test proposed changes against a digital twin of their production network before deploying them.

The company says Predict runs every proposed change against a mathematically accurate model of the entire production environment, flagging outages, security exposures and policy conflicts before they reach live infrastructure. Forward is pitching the product as the foundation for what it calls autonomous networking, in which artificial intelligence agents propose, validate and execute changes without human intervention.

The release coincides with a corporate rebrand from Forward Networks to simply Forward. The product will be available in fall.

The release is seeking to address the issue whereby, for years, network engineering has lacked an equivalent of the staging environments that software developers rely on. Change windows are planned weeks in advance, run through method-of-procedure documents and Change Advisory Boards and still routinely produce outages because the production network has been the only true test environment.

Forward Predict aims to close that gap by validating proposed configuration changes against a twin that spans every vendor device and major cloud, including Amazon Web Services Inc., Microsoft Corp.’s Azure, Google LLC’s Cloud and IBM Corp.

The underlying digital twin models device state from the network layer through the application layer and identifies all possible packet paths and policy conflicts. Predict extends that capability to future network states, returning deterministic evidence of how a change will behave once deployed. When a change fails verification, the platform returns specific failure data, which Forward says allows AI agents to iterate until a fully verified configuration is produced.

“Forward Predict changes the paradigm by verifying every change before it happens, which means engineering teams stop operating on instinct and start operating on certainty,” said co-founder and Chief Artificial Intelligence Officer Nikhil Handigol. “This level of certainty is the essential prerequisite for autonomous networking, enabling AI agents to propose, verify and deploy at machine speed without human intervention.”

The product is in beta with a group of customers. Steve Bamford, a senior network engineer at IG Group plc, said that he had deliberately staged changes that would have isolated parts of the network if pushed live and that Predict caught them.

Forward’s customer list includes Goldman Sachs Group Inc., PayPal Holdings Inc., S&P Global Inc., IBM and Dell Technologies Inc. The company cites an International Data Corp. report claiming the average Forward customer realizes $14.2 million in annual benefits from efficiency and security gains.

Forward is a venture capital-backed company that has raised $116 million in funding, including $16 million in August 2017 and $50 million in January 2023. Investors in the company include Andreessen Horowitz, Goldman Sachs, MSD Partners, Section 32, Threshold Ventures, A. Capital and Omega Venture Partners.

Image: Forward