Artificial intelligence has a wide range of applications and use cases. But a recent article by Kailem Anderson suggests that traditional telcos have the most to gain from leveraging AI.
“Much of what service providers and telcos do involve manual tasks that take up the majority of a network technician’s day,” says Anderson, vice president of software and services at Ciena. “AI will help these providers improve their operational efficiency and provide a much better and tailored customer experience.”
Indeed, AT&T (News - Alert) is already working to advance and adopt AI on a number of fronts.
“In early 2015, we stated our plan to virtualize and software-control 75% of our core network functions by 2020. Last year, we hit our goal of reaching 55%,” Chris Rice, senior vice president at AT&T Labs (News - Alert), Domain 2.0 Architecture and Design, wrote in a March blog. “And we just announced that our goal for 2018 is 65%, which we intend to hit too. But this latest goal is particularly critical, as 2018 is also the year we plan to launch mobile 5G here in the U.S.”
AT&T is also working with Tech Mahindra on an open source platform and marketplace called Acumos. The platform enables data scientists to publish adaptive AI models without the need to develop fully integrated solutions.
“It packages each model into an independent, containerized microservice, which is fully interoperable with any other Acumos microservice, regardless of whether it was built with TensorFlow, SciKit Learn, RCloud, H2O or any other supported toolkit,” Rice explains. “Models built with any of these tools or any supported language, including Java, Python and R, can be automatically onboarded, packaged and cataloged. These microservices are easy to integrate into practical applications, for any software developer, even without a background in data science or knowledge of any specialized AI development tools.”
Edited by Maurice Nagle