MWC 2024: Revolutionizing Telecom with Generative AI
MWC 2024: Revolutionizing Telecom with Generative AI
GenAI holds the power to introduce transformative change in the communications industry – from simplifying the operations of networks to streamlining interactions with customers, and more. Amdocs’ Anthony Goonetilleke shares his perspective.
I was recently reminded of an old Latin adage, ‘Tempora mutantur, nos et mutamur in illis,’ which translates to “Times change, and we change with them.”
The communications industry has been a vital part of every technological change, from the first telephone call in March 1876, to major shifts like the internet, the dot com era, the iPhone, cloud, and even catastrophic events like COVID. Generative Artificial Intelligence (GenAI) is yet another technological paradigm shift that will change the way we work, play, and innovate. Communication service providers (CSPs) all over the world are ready to embrace it, but it isn’t without challenges and complexities.
The business impact of GenAI
Over the past year, I’ve traveled the world to chat with c-suite executives from some of the largest connectivity businesses and our key strategic partners, including Webscale giants and those at the forefront of accelerating Large Language Models (LLMs). The message is clear: the business impact of GenAI will be significant. However, it’s essential to deliver carrier-grade service levels and trusted AI that global brands demand while maintaining high accuracy levels. This is vital to embracing and monetizing the technology.
For one, establishing robust guardrails and utilizing telco-specific GenAI models is essential to prevent potential hallucinations, which can be common in consumer-grade GenAI. Further, concerns such as bias and ensuring observability are also key considerations, as well as GenAI providing varied responses to identical inquiries, especially when it comes to consistency in pricing, are simple examples of the complexities of large-scale implementations.
"GenAI is a game-changing force. Organizations and industries must reconsider everything from the talent they are nurturing and the tools they are employing, to the objectives they are setting, and concentrate on how GenAI can best serve their people, organization, and customers."
Considerations before GenAI implementation
As with any technology change, workforce training, governance, security, observability, and scalability are critical components that must be addressed. And of course, GenAI’s results are only as good as the data you are feeding it.
Data readiness underpins not only generative AI, but more broadly, the ability for telecoms and other enterprises to tackle AI and even automation. It is why we see such a broad range of implementations. Telco data is complex, and the underlying systems are often a combination of different architectures, technologies and so on, making data challenging to ingest.
Before CSPs start the generative AI impact discussion, they need to ensure they have a clear strategy for their data, a robust approach to governance, a well-defined architecture and, if needed, a plan for migration and consolidation. Prioritizing and creating a viable data ingestion pyramid becomes one of the most important infrastructure considerations for a foundational long-term strategy.
Creating a telco native GenAI platform
GenAI holds the power to introduce transformative change in the communications industry – from simplifying the operations of networks to streamlining interactions with customers, and more.
At Amdocs, we are investing heavily in ensuring that CSPs can rapidly and safely adopt GenAI in every aspect of their business. We understand the importance of striking the right balance between innovation, responsibility and being cost efficient.
To ensure CSPs can leverage this transformative technology, we created a GenAI platform, amAIz, which brings new GenAI-powered applications to customers faster, enabling them to benefit from the immense potential of GenAI, while also providing enterprise-grade security, reliability, and performance.
By leveraging the best of the industry’s evolving foundational large language models (LLMs) and overlaying a robust vertical taxonomy and pre-configured telco-specific ‘use case kits,’ CSPs simply need to choose what generative AI capability they would like, and the underlying technology integration, training and governance is handled.
With any major technology shift, ecosystems and partners are important as the technology landscape both accelerates and fragments. amAIz is built in strong collaboration with industry leaders. In partnership with Microsoft, we have built Customer Engagement Platform , which includes multiple copilot products, including sales, order enrichment, care, proposals and more, providing an interconnected experience beginning to end. We are also collaborating with NVIDIA on optimizing large language models (LLMs) to improve latency, accuracy and cost effectiveness of generative AI applications and services. These partnerships are essential as we transform the industry.
The importance of telco-specific copilots
As GenAI becomes more embedded across various aspects of the business, it is critical to use GenAI copilots that are fine-tuned with telco data sets and telco taxonomy, and leverage RAG (Retrieval Augmented Generation) to improve accuracy and optimize the costs of tokens. We embed an architecture of vector databases to augment the flow of information to drive cost effective and accurate results. In recent benchmarks the amAIz platform has shown an improvement of up to 30% on standard public LLM architecture.
With this, we have embedded AI-driven assistants (copilots) across our portfolio, allowing CSPs to augment human capabilities and creativity across domains. For instance, a copilot that can help businesses create proposals and optimize pricing, AI-driven insights that can guide recommendations on new offerings for consumers in near real-time, and others that pinpoint network events that are impacting customers.
The promise of GenAI
In our early work with customers using amAIz, we’ve seen proof points of the promise that GenAI holds for telecom and beyond. Robust customer care, commerce and experience data is enabling us to derive unique business insights and further leverage generative AI to create new outcomes for operators and their customers.
But what excites me the most is that this is just the beginning. Companies like Meta, OpenAI, NVIDIA, Microsoft, and others, are making tremendous progress on defining what comes next – and we are thrilled to help bring this to the communications industry. The collaborative spirit in the AI community is propelling us forward at an unprecedented pace.
GenAI is a game-changing force. Organizations and industries must reconsider everything from the talent they are nurturing and the tools they are employing, to the objectives they are setting, and concentrate on how GenAI can best serve their people, organization, and customers. We believe in business-led initiatives and focus heavily on ROI models for all our projects. The future is bright for GenAI, and you are either embracing it or have been left behind.
This article was original written for the GSMA Show Daily at Mobile World Congress 2024
