A company specialising in digital twin technology has told TMN that it has helped one mobile operator save outage costs, increase speed of deployment and reduce downtime by applying its technology to live network operations.
Forward Networks, which has today announced Forward AI, said that its technology gives operations teams the ability to ask complex questions and validate outcomes of automated processes in natural language by accessing a “mathematically accurate” digital twin of the network built on “trusted data”. Forward AI will be included in the company’s Forward Enterprise platform and is scheduled for general availability in April 2026.
Operator outcomes
Although the company was not able to name the operators that gained results, TMN understands Forward Networks has had engagements with Vodafone and Verizon, amongst others. A spokesperson told us, “One mobile operator accelerated its deployment velocity by 70% and reduce unplanned downtime by 33%. In another testament to the benefits of a network digital twin for telecom operators, a Forward Networks customer calculated $11M in savings by increasing the efficiency and eliminating outages related to change management and automation only.”
“When we founded Forward Networks, we started with a simple but strategic question: does the network actually behave as intended, by design, in production, and across changes,” said David Erickson, CEO and co-founder of Forward Networks. “Answering that question requires more than visibility. It requires a mathematically accurate model of the network itself.
“That is what Forward Enterprise delivers. Forward AI builds on that foundation to allow enterprises to reason about their networks, validate outcomes, and take action with confidence.”
Nikhil Handigol, co-founder and chief AI officer at Forward Networks said, “Forward AI extends the power of a mathematically accurate digital twin across operations, enabling organisations to safely adopt AI-driven workflows, reason transparently about outcomes, and preserve the safeguards required to operate critical infrastructure.”
Forward AI says that by creating a behaviorally accurate understanding of the network the digital twin model establishes a foundation for the safe adoption of agentic AI in network operations, delivering validated, evidence-backed recommendations. The company added that it would support the Model Context Protocol (MCP), making its verification foundation available to enterprises and third-party developers, enabling them to build agentic tools grounded in accurate, current network data.
“This is how AI becomes a trusted part of operating critical infrastructure,” Erickson said.
The contextualisation and validation of data, application of knowledge graphs and digital twin capabilities are seen as key to the introduction of advanced AI and agentic AI capabilities in network operations. Most operators are grappling with how to advance in this area, seeing potential efficiency and monetisation benefits.
A blog post from Forward Networks’ Nikhil Handigol, Co-Founder, claimed:
“Forward AI is not just another thin wrapper around an LLM. It is a conversational agentic system designed to plan, reason, and execute multi-step workflows using mathematically validated network data, built natively on the industry’s first mathematically accurate network digital twin. While general-purpose AI often relies on assumptions, Forward AI acts on trusted data.
“When you submit a prompt, Forward AI doesn’t approximate an answer based on assumptions or partial context. Instead, it constructs an execution plan, performs deep network analysis, and produces results accompanied by clear explanations of the underlying reasoning. Every outcome is derived from domain-specific evidence, allowing operators to understand not just what the answer is, but why it is correct.”
aifrisuren
This is fascinating. The application of digital twins to live mobile network operations, especially to reduce downtime and save outage costs, seems like a huge leap for operator efficiency. I wonder how complex the initial data ingestion and model validation process is for a live, high-stakes environment like that.