Generative AI (artificial intelligence) models like ChatGPT have grabbed the world's attention over the last year. In response, many applications of generative AI have been appearing across almost every industry. As a technology leader, the telecommunications sector, particularly the Operations Support Systems (OSS) and Business Support Systems (BSS) industry, is no exception.

This article is the first in a two-part series about the different modes of artificial intelligence that are applicable to network operators, as well as some of the potential use-cases. This first article delves into the fascinating realm of generative AI and its transformative impact on OSS/BSS. The second article looks at discriminative AI.

How generative models like ChatGPT are revolutionising OSS/BSS in telco

Generative models aim to understand the underlying distribution of data and patterns within large data sets. Generative models do just that. They analyse and uncover patterns and relationships in a data set to then generate new samples that resemble the observed data. It can quickly generate plausible text and other data based on what it has learned from the base data. Given the huge amounts of data collected by and generated by telcos every day, they have a lot of data from which generative AI models can learn.

Unleashing the power of generative AI: Eight examples of its use in telecommunications

Use Cases in OSS/BSS and Telco:

  1. Document Generation: Generative models can assist many different types of workers who interact with OSS and BSS tools with creation of documentation. This ranges from outlines for network or architecture designs through to summarising knowledge-base entries from a set of fault-fix activities that had previously been performed
  2. Network Topology Design: By understanding the underlying relationships between different network elements, both new and existing, generative models can suggest optimal designs, placement of assets and connections 
  3. Network Traffic Simulation and Test Data: Generative models can create realistic network data simulations. These can be used for a variety of purposes including network planning, training simulations, helping in capacity optimisation and performance enhancement as well as generating synthetic data to support manual or regression test scenarios
  4. Customer Behaviour Modelling: Generative algorithms can create customer profiles and simulate realistic usage patterns for these profiles. This data can be used for multiple purposes including testing, network / service optimisation and even for facilitating personalised marketing strategies
  5. What-if Analysis: By generating synthetic data relating to networks, services, traffic and customer behaviours, as mentioned above, generative models can be used to explore multiple hypotheses, configurations or scenarios that help with optimisation of the network
  6. Exploratory Analysis on Log Data: Log data can be a rich form of insight, but often goes underutilised because of the challenge associated with processing unstructured data (i.e. the text format of logs). Generative models provide a nuanced way to explore relationships and patterns in log data, offering insights that help to unlock operational efficiencies
  7. Customer Support: Generative AI can create dynamic feedback and support content, adapting to customer queries in real-time. Collecting and assembling customer data in near real-time can provide a more personalised service experience.
  8. Personalised Service Bundling: By understanding the diverse needs of customers, generative AI models can generate service bundle combinations that uniquely align most closely with their specific requirements and preferences

From chatGPT to network simulations: the impact of generative AI on telecommunications

Generative AI techniques like these are revolutionary for telcos in the following ways:

  1. Efficiency: The examples listed above had all been time-consuming to prepare, often very manually, in the past. Generative models not only make these use-cases accessible, but also make it possible to perform them in more sophisticated and nuanced ways
  2. Broader / Deeper Insights: The more sophisticated and nuanced use-cases allow generative models to unlock an even deeper understanding of customer behaviours, network patterns, data relationships, etc and in doing so can enable more proactive and refined strategies / responses
  3. Innovation and Creativity: These models open up entirely new avenues for product development, user-interfaces, testing, automation, optimisation and enhancement.
  4. Complementing Discriminative Models: In some applications, generative models can work hand-in-hand with discriminative models, enhancing the overall performance and capabilities of OSS/BSS systems, as we’ll describe in the next article.

Transforming telecom with generative AI

As you may have already perceived, Generative AI has the potential to revolutionise the telecommunications industry, particularly within our field of OSS/BSS systems. From traffic simulations to customer behaviours / engagement, the possibilities are vast and transformative. Stay tuned for the second part of this series, where we will explore the world of discriminative AI and its applications in the telecommunications industry.