-
Chatbots dominated the telco AI landscape last year, but horizons have broadened
-
Google Cloud's Ankur Jain said telcos are now looking at use cases spanning network operations, field operations and sales and marketing.
-
The focus in 2024 is less on experimentation and more on how to scale responsibly
After a year spent dating around and testing out different artificial intelligence (AI) applications, it seems telcos are finally ready to settle down with the good ones.
Last year, Google Cloud’s telco chief Ankur Jain told Fierce Network operators were mainly experimenting with AI for customer care and IT operations. But during a fresh interview this month, he said their horizons have broadened a bit.
According to Jain, there are four primary categories of AI apps telcos are adopting. Customer care remains one of these, but operators aren’t just using AI for chatbots. Instead, players like Verizon are also using AI to help surface relevant information for their call center reps and summarize calls to allow reps to move on the next customer faster.
On the network side of the house, AI is still being used to identify, diagnose and remediate issues faster. But it’s also now being applied by operators like Telefonica to analyze power usage to improve energy efficiency, Jain said. There are also more applications targeting anomaly prediction to address issues before they even occur.
The third bucket of AI applications focuses on field operations, Jain said, with AI helping operators reduce truck rolls. How? Well, he said technicians can submit a photo to an AI system to immediately identify a problem – avoiding duplicate trips. This is key in an industry where the capabilities of field ops teams vary.
“In some cases they are much better trained and in others they are outsourced and they may not know how to fix an issue,” Jain explained. “By providing a multi-modal capability [that can process photos]…they’re able to detect what the issue is and it’s able to suggest what the fix would be.”
Jain said there are “multiple operators” who are working with it on applications in this realm.
The final bucket of AI applications is in sales and marketing. Jain said while this category is popular with telcos, Google is seeing adoption on this front across verticals. In this space, AI can help generate better marketing materials, suggest “next best offers” or provide more customized recommendations for customers, he said.
Jain’s assertions are backed up by data from a recent survey of 400 telecommunications players conducted by Nvidia. That survey found 43% of respondents said they are investing in generative AI. Just under a third of respondents said they invested in at least six AI use cases last year and 40% said they’re planning to press ahead with scaling six or more use cases in 2024.
The top use cases telcos said they were adopting included improving customer service and support (57%), improving employee productivity (57%), boosting network operations (48%), network planning and design (40%) and generating marketing content (32%).
Plugging it into a platform
As part of the shift from experimentation to implementation, Verizon SVP and chief data officer Kalyani Sekar noted during Google Cloud Next earlier this month the operator has homed in on developing responsible AI practices. Other executives on the panel she took part in agreed: governance is the industry’s current watch word.
Jain indicated the focus on governance is proving to be a boon for the platform strategy peddled by Google Cloud and other major players.
“They’re all beginning to come to the realization or have come to the realization that they need a platform,” he said.
“They need a common platform that they can use across their company so that they have common governance models, they have a common model around data provenance, around making side-by-side quality experiments or managing the full lifecycle of new models that are coming in,” he concluded.