Mavenir pitches GenAI for network troubleshooting

Called Operations Co-Pilot for RAN Service Assurance, it claims to be able to automate network troubleshooting, giving telcos the means to anticipate issues and fix them before they have a chance to become a major problem.

Mavenir has trained large language models (LLMs) with detailed KPIs, counters, logs and traces from network infrastructure, giving its new offering a deep well of knowledge to draw from.

The new framework is built on Nvidia Tensor Core GPUs and the Nvidia AI Enterprise software platform for GenAI, and the whole thing is hosted on AWS.

Telcos for years have been talking about the benefits of using AI/machine learning (AI/ML) to enable automated RAN operations that can handle increasing network complexity.

Benefits include zero-touch optimisation, managing energy consumption, reducing human error, and generally making networks software-led and programmable. These – and others – are routinely touted by vendors pitching their intelligent RAN solutions.

When it comes to GenAI for telcos, much of the discussion to date has centred on leveraging chatbots to improve and automate marketing, customer service and back office processes, improving efficiency and freeing up resources to be put to use elsewhere.

McKinsey predicted back in February that GenAI could add $100 billion of incremental value to the telecoms industry, on top of $140 billion-$180 billion in productivity gains.

Now Mavenir, together with Nvidia and AWS wants to put GenAI to work deeper in the network.

"Our new Operations Co-Pilot framework has the potential to be a game-changer for operators, delivering a wealth of fault prediction and root cause analysis capabilities with AI-powered accuracy and speed," said Mavenir's chief technology and strategy officer Bejoy Pankajakshan.

"The three key use cases of Core Dump Analysis, Log Similarity Search feature and Log Anomaly Detection can be used together to maximise the automated efficiency gains," he continued. "Combining information from these trained LLMs will provide broader and deeper coverage of all fault and anomaly scenarios, enabling very early prediction of potential problems in the network. Moreover, as these solutions evolve, the models have the ability to continuously learn and optimise based on user feedback and iterative training with further logs, KPIs and traces, promising even greater gains."

The announcement serves up further evidence of the impact that GenAI is having on hyperscalers like AWS. The latest prediction – courtesy of IDC – is that thanks largely to GenAI, public cloud spending will grow to $800 billion this year from $669.2 billion last year.

It's also yet another piece of good news for Nvidia. Its GPUs are proving a hit with GenAI service providers – which includes an increasing number of telcos. It is little wonder that Gartner predicted last month that AI chip revenues are expected to jump 33% this year to $71 billion.

"In today's fast-paced digital landscape, generative AI is a force multiplier for enterprises everywhere," said Chris Penrose, global head of business development for telco at Nvidia. "Mavenir's framework, built on the Nvidia AI Enterprise platform for streamlining the development and deployment of AI, will unlock a new level of operational efficiency for communications service providers."