As anticipated, artificial intelligence (AI) and its impact on telecom was a key topic at this year’s DTW Ignite in Copenhagen. Mavenir’s Senior Vice President and General Manager for its Digital Business Enablement Sandeep Singh spoke with RCR Wireless News about how the company is tapping into the potential of this technology to support its communication service provider (CSP) customers.
First, Singh shared, Mavenir is helping CSPs tackle operational pain points by leveraging AI to gather much more data than previously possible. This data, he explained, comes from networks logs and statistics. “We are using that data to train the models to bring anomaly detection and to solve a lot of operational problems for the operator,” he continued.
The company is also capturing customer information and then using AI logic to offer “next best products” tailed to each customer based on their past behavior and usage. According to Singh, this is a “huge benefit” for CSPs as it will enable them to better compete in the marketplace.
Another necessity to remain competitive, claimed Singh, is getting products and software on the market as fast as possible. “We are using AI to build new … high-quality software … which reduces total time of development,” he said.
Finally, Mavenir is using AI to expose the network using APIs. “We are talking to operators about how to take the network capabilities set and expose this to third party developers and other partners,” Singh explained, adding that doing so, will allow them to not only become the enabler for those partners, but also benefit from the resulting new revenue line. “We are able to not just expose that network function, but also help operators monetize and implement new commercial models,” he added.
Further, the use of APIs allows Mavenir to execute on the data in a “very surgical fashion.” Singh provided the example of being able to more precisely and quickly train a network parameter to reduce future critical anomalies in the network.
“At the end of the day, AI is trying to solve where we are putting a lot of human power … so a lot of operational excellence can be brought in,” concluded Singh. “Every decision point in the network, we are able to extract intelligence from … then we are able to clean and increase the quality of the data … [and] the AI foundational models on which we are training [is] getting better [through] technological advances.
