Around the world, major telecom operators are leveraging the latest advances in artificial intelligence to better manage their customer relationships or help deploy new high value-added services.
Like all business sectors, the telecom industry is also counting on taking advantage of all the opportunities opened up by Artificial Intelligence (AI). All market players, be it consumer and B2B operators, manufacturers and engineering companies, are moving to integrate AI into their activities.
From your point of view, what is the added value of Artificial Intelligence for telcos?
The ability of artificial intelligence to shape and solve the most critical problems is creating new opportunities in the telecom industry to improve customer experience and operational efficiency across teams and networks.
Provide a new dimension to customers' relationships
AI is at the heart of transformation programs to improve customer management, while optimizing the added value delivered by consultants.
- Agility and efficiency: When we think of AI and customer relations, we think first of chatbots or callbots. Callbots are directly accessible by customers 24/7 and allow to qualify and potentially process customer requests, whether they are commercial or technical.
- Prediction and anticipation: Artificial intelligence can also be used to improve the personalized services offered to customers, to anticipate their needs and optimize cross-selling and upselling. Customer interactions can be digital via the operator's portals and mobile applications that integrate these predictive capabilities, or via advisors in stores and call centers who then have access to predictive tools. By identifying and anticipating potential needs or friction points, AI becomes an excellent tool to reinforce customer satisfaction and loyalty.
Improve network management and service quality
At a time when networks are becoming increasingly complex (increasing number of services, diversity of providers, development of virtualization, ...), Artificial Intelligence can provide crucial help in managing networks and improving the quality of their services. Thanks to the very large amount of data provided by telecommunication networks on their operations, and thanks to predictive models from AI, many use cases of AI in network management have arose. These include:
- Predictive maintenance: Some technical failures (e.g. unavailability of resources, VPN malfunction, etc.) can already be identified or even anticipated. Anticipation is based on data sets describing previous malfunctions and knowing the user impacts.
- Crisis response: Once an error is detected or anticipated, AI can quickly assess all possible solutions, can simulate their deployment and measure their effects, in order to propose them to a human operator who can then make a choice among the proposed solutions.
- Optimizing network operation in real-time: AI can also consider all the changes it can make to the network configuration and simulate the deployment of these changes in order to evaluate their effects. Artificial intelligence can thus be used to optimize network energy consumption by putting equipment (antennas, servers, fibers, etc.) that is unlikely to be used on standby or, on the contrary, by dynamically allocating more resources to them.
- Network capex optimization: Some services rely on many metrics that can affect the quality of the user experience. Many of these parameters must be considered as a whole to measure their effects. Without knowledge of these parameter dependencies, it can be very difficult for a network engineer to optimize network parameters for a given customer, but this can be done with an AI tool.
There are many other uses of AI in networks (quality of service optimization, fraud detection, etc.), and this field remains to be further developed. To do so, more data will have to be collected and new predictive patterns will have to be built.
Orange is multiplying studies and projects in order to optimize network management and the quality of their services.
Improve team efficiencies and performance
The AI use cases described above will de facto improve the productivity of customer facing teams (call centers, stores), and those in charge of network maintenance.
This optimized productivity thanks to AI can also be found in many other activities; we will focus hereafter on the activities of the Fiber technicians.
Fiber deployments are a priority for operators. Relying on artificial intelligence to assist technicians in the fiber rollout not only improves their productivity, but also optimizes the quality of the fiber network.
For example, an AI application can allow the technician to take pictures of the key elements of his worksite (for example, take a picture of the fiber entry point or the optical termination point). The AI analyzes each photo taken by the technicians, it instantly recognizes a non-conformity, dialogues with the technician and allows him to repair the problem online. Technicians thus avoid having to re-intervene due to the non-conformity of the site, which generates additional delays, non-satisfaction, as well as additional operational costs.
AI, acting as a personal assistant for each technician, allows to connect customers faster, to optimize costs, and to be sure that 100% of the works are checked.
What are the challenges of artificial intelligence?
The adoption of AI in business enables working on ever wider and richer scopes of work. It offers a new vision and innovative perspectives to rethink business processes and customer relations and thus gain in performance and productivity. This transformation implies numerous changes at all levels of the organization.
The first challenge of AI is a competencies challenge. Its deployment requires the acquisition of new capabilities to master new technologies, and the emergence of new jobs. The training of existing teams and the recruitment of new skills are crucial.
The second challenge is organization: How to ensure that AI use cases can be developed in the enterprise's various activities? How to get business teams and AI teams to work together? How to gather and structure all the technical, commercial and marketing data that exists within the enterprise? AI is applicable everywhere, but should we centralize AI experts and resources, or distribute them among the operational teams?
Each organization may have its own answers, but it is important that their questions are well addressed.
The third challenge is related to trust and ethical responsibility: AI relies on the handling of large datasets, for learning (machine learning / deep learning algorithms). Therefore, it makes operators accountable for using this data, and for respecting privacy and individual rights.
It also involves issues of transparency and fairness: the aim is to guarantee that it’s learning algorithms do not generate inequalities or exclusion.
These challenges must be addressed communally to make AI a major factor of development for customers and the company.
How does Sofrecom support this transformation?
Sofrecom, a consulting and engineering company specialized in the Telco sector, supports telecom operators in all transformations, and of course in those induced by artificial intelligence.
Our expertise in AI is based on 4 main issues:
- Data Transformation: We co-design with our customers the best organization to operate the Data foundation that is needed for AI programs,
- Optimized network rollout planning: We optimize investment choices in fixed, mobile and physical distribution networks,
- Tailored marketing and customer experience: We help develop customized and differentiated customer experiences;
- Integrating AI solutions: We help our clients to select high value-added AI solutions for their specific needs, and to implement them in their ecosystem.
These are major challenges for Telcos, and their ability to leverage the potential of AI in their business will be key to their success.
