At a time when excitement in artificial intelligence (AI) is reaching fever pitch, communications service providers (CSPs) are making their first tentative forays into a field of computer science which has the potential to bring significant benefits to their businesses.
In this newly published report, TM Forum has identified five categories of use cases for AI in CSP businesses:
- Service management
- New products and services
- General business functions
Most activity today is in customer-centric solutions, more specifically chatbots. A TM Forum survey conducted in September and October this year, three out of 10 CSP respondents said that they had rolled out chatbots. Even though CSPs often measure the success of mobile apps and chatbots in terms of the reduction in the number of calls to call centers, many respondents said that the key benefit from the introduction of chatbots was that it allowed their call center staff to spend more time on higher value tasks such as cross-selling and upselling.
Service management is a focus for many CSPs today. The deployment of AI solutions to, for example, prevent outages and take a proactive approach to operations more generally, is on many CSPs’ roadmaps although vendor-supplied solutions are only now beginning to emerge.
Network, and more specifically, network automation is, arguably, the most exciting potential use case for CSPs. While SDN and NFV clearly represent the future of the telecoms network, making that transition without incurring significant short- to medium-term cost increase is a major challenge and inhibitor. AI has the potential to achieve radical automation and, in doing so, to lower costs. In the longer term, AI can enable operators to take more of a ‘top-down’ approach to designing networks that meet specific business objective and goals.
We are still at the experimentation phase in AI today and CSPS are trying to figure out how best to structure AI initiatives within their organisations. The most likely approach – and one pioneered for example by Verizon in the US – will see AI initiatives being undertaken across the organisation with different vendor partners and solutions but the creation of a centralized team that manages the data and explores potential synergies and collaborations.