As telecoms networks become increasingly complex, CSPs need to employ advanced AI tools to maximize service uptime and enhance customer experience. To that end, this Catalyst is exploring how CSPs can harness generative artificial intelligence (genAI), large language models (LLMs) and digital twins of their networks to improve the efficiency of field maintenance engineering (FME) operations. In particular, the Catalyst is developing an AI assistant for home broadband and mobile broadband FME. It is also using a digital twin to help assess and prioritize service impact and support CSPs’ network operations centers (NOCs). To assist an engineer working in the mobile broadband back office, a genAI agent can automatically analyze a fault’s service impact, root cause and EDNS (expected demand not served) metrics using ‘chain of thought’ reasoning. In the home broadband FME domain, a genAI agent can assist engineers by conducting service troubleshooting, testing and provisioning. The output of the Catalyst is designed to provide a reference for the telecoms industry on how to best make use of genAI, LLM-based co-pilots and digital twins to simplify their network operations. The Catalyst is harnessing the work of multiple standards organizations, including the TM Forum’s frameworks for autonomous operations and networks, TM Forum’s Autonomous Operations Maturity Model (AOMM), and ETSI’s Zero-touch Service Management (ZSM) project. The Catalyst team is employing an evaluate, operate, transfer (EOT) process and value metrics to assess the likely return on investment of genAI solutions. The first step is aligning with the CSP’s C level management to confirm the objectives. The CSP and vendor will then take a top-down approach to establish a key business index, key performance indicators and operations performance metrics. The Catalyst intends to evaluate its impact by measuring the impact on churn, cost savings, revenue and complaints. In a proof of concept, traffic loss was reduced by between 7% and 15%, which could increase a CSP’s revenue by millions of US dollars annually.