Zephyrnet Logo

Driving connected personalised user experiences with Generative AI | IoT Now News & Reports

Date:

As the world continues to rapidly move towards digitalisation, customer expectations are also on the rise. Around the globe, telcos are grappling with meeting these expectations. As well as ensuring connectivity in a secure, seamless, and consistent manner 24/7, to compete and differentiate, operators now need to provide personalised experiences that are as unique as the customers themselves.

With customer information including usage patterns from billing and business assurance systems, interaction profiles and usage behaviour from CRM and fraud management systems, and geo location and customer activity from deep packet inspection, the raw data on customer behaviours and preferences is already available. The challenge, however, has always centred around how to analyse the data in a timescale that makes it powerful, usable, and relevant in terms of actionable insights that can be utilised in near real time to generate truly personalised services.

The key to unlocking this new personalised world lies in the advent of Generative Artificial Intelligence – or GenAI. Unlike ‘conventional’ AI, which is more like a classifier in specific domain use cases, GenAI learns from a wide range of knowledge and uses this to create an arguably infinite range of content. GenAI utilises advanced algorithms to create new content based on given prompts, which can be in the form of images, text, 3D models and music.

GenAI is having a major impact on the telco industry and is beginning to change the way we communicate. It is already being utilised to reshape the way services are provided, networks are managed, and customer experiences are enhanced.

The evolution of personalisation

The power of GenAI lies in its speed and depth of learning, and this is why it is becoming so valuable in the telco space. Think of customer segmentation – maybe operators do that on a regular basis, or maybe not so regular. Either way, the results are always out of date, making targeted advertising, offers and support more generalised than personalised.

With GenAI’s speed of understanding, customer segmentation can be dynamic – recommendations and offers can be pushed to the customer in real time. With its ability to call on knowledge, functions, tools & skills and inherent flexibility, GenAI is leading the evolution of personalisation. We have all received messages and offers that are not relevant to us – with GenAI, customised advertisements and promotions can be generated, making the customer feel that their service provider is making the effort to treat them as an individual.

Take the example of music and video. A customer is watching football on YouTube but does not use YouTube for music – for that they choose Spotify. GenAI recognises that difference instantly and can generate totally relevant offer in real time for the customer – a bigger data bundle, an upgraded handset with a larger screen, discount offers or early access to tickets to sporting and music events – all to the customer’s known interests and taste.

Dramatic rise in first call resolution

The customer interaction with service and support options can also be significantly improved with the integration of GenAI. The increasing complexity and range of telco products, offers, bundles and promotions make it almost impossible for call centres and chatbots to have a full understanding of each. This means there are limitations on the efficiency of these real and virtual assistants in handling natural language queries and conversations.

Billing issues are one of the largest causes for calls to helpdesks, and customers can become irate quickly if their concerns are not immediately understood and resolved. First call resolution is low, customers are unhappy, and revenue assurance becomes threatened. GenAI can understand the intent of the customer – if the customer is getting angry during a conversation, GenAI can see that the interaction is not going the right way, understand what the problem is and prompt the agent to give the correct reply by automatically pulling the right data and policy knowledge – that’s first call resolution. We have seen up to 30X ROI in in this use case once GenAI is integrated into our client networks. It’s too early to be quoting return on investment figures, but the accuracy of response is increasing dramatically.

GenAI Adoption is easy but needs a fine tunes approach

It is of course necessary to ensure the GenAI model is properly trained – these Large Language Models (LLMs) have a huge wealth of knowledge, but it will not initially be telco-specific. There is a skill to making GenAI more telco-lingo driven to ensure it is fully aligned with the business. It is also essential to have guardrails in place to prevent unintended consequences – a chatbot going ‘off script’ for example. And of course, security always has to be top of mind to avoid malicious content being introduced – bad actors injecting prompts, code and substrings. Guardrails need to be built to specifically prevent this.

Integration of GenAI into existing infrastructure is not as complicated as you might think. The most powerful models work best in the cloud, but telcos do have to follow legal and regulatory frameworks, and this could mean keeping the data on-premise. The choice is of course with the telco. Either way, GenAI can play a central role in transforming purchasing and support interactions into personalised experiences. The arrival of GenAI into the telco world is re-energising the sector and revolutionising the way services are provided. Embracing this innovation will be critical to those operators looking to exploit the full potential of the digital world.

Harsha Angeri, VP, Corporate Strategy & Head, AI Business, SubexHarsha Angeri, VP, Corporate Strategy & Head, AI Business, Subex

Article by Harsha Angeri, a VP of corporate strategy and head of AI business, at Subex

Comment on this article below or via X: @IoTNow_

spot_img

Latest Intelligence

spot_img