The agentic AI motor can transform CX, with the right chassis

Growth has stalled and customer satisfaction has plummeted. Net Promoter Score (NPS) languishes at around 31, placing telecoms among the lowest-ranked industries for customer experience.  

It’s not just a reputational concern - fixing it is a strategic imperative.  

Because, in an era where digital-native brands have redefined consumer expectations, the traditional telco model - reactive, segmented, and slow - is no longer fit for purpose.

Telecoms face a clear choice. They can continue to compete on price, risking further commoditisation (and ruin) or they can evolve into smart, proactive service providers that customers genuinely value.

‍

The CX Gap

The nut of the problem is in how telcos treat their customers.  

For years they have relied on broad segmentation and batch-data processing - and the IT infrastructures built to handle it - to manage their relationships.  

It’s an approach that treats customers as statistical averages rather than individuals.

By the time a churn signal is detected, the customer has often already mentally checked out.  

That’s because customers crave relevance, immediacy, personalisation and anticipation of their needs.

Responding to this shift in expectations is a growth imperative for a stagnating industry. It’s also a huge customer experience (CX) challenge.  

‍

GenAI beyond the buzz: what really matters for telcos

“AI agents” are everywhere in conversations about telecom innovation, but so far, most so-called solutions are just basic LLM workflows or RAG implementations – far from the autonomous systems promised.

‍

GenAI is an evolution, not a single tool:

• LLM Workflows – simple text generation, no reasoning.

• RAG – adds accuracy by pulling data from trusted sources.

• AI Agents – reason, plan, and automate tasks.

• Agentic AI – multiple agents collaborating for end-to-end automation.
• ‍

Each layer solves a different problem. The key is using the right one where it delivers real value.

‍

triPica’s GenAI Journey: Practical, Not Just Promises

triPica has always been at the forefront of telco innovation, naturally we were among first to be seeking how we can turn AI into practical solutions for our customers, reliably. We demonstrated the potential of LLM and RAG for telecom in 2024 demonstrating live at the DTW real case scenarios showing how telcos can boost customer service efficiency and optimise customer experience.  

Now, we’re evolving further, building Agentic AI capabilities using AWS Bedrock to equip telecom operators and MVNOs with AI that truly helps agents work faster and smarter – as seen with our AI Account Summary tool.
‍

‍

While Agentic AI has been hyped as the magic answer, we see few pilots making it into production – too costly, too complex, too error-prone. triPica’s approach is pragmatic: deploy AI where it drives tangible efficiency, not just buzzwords. Our focus is clear: AI that drives measurable efficiency today while preparing for tomorrow.

‍

Agentic AI: A panacea for broken CX?  

Let’s dwell deeper on the Agentic AI - a class of artificial intelligence that perceives context, makes decisions, and acts autonomously across systems - can be part (but not all) of the answer.  

Agentic AI has the potential to shift from reactive service to proactive engagement.  

It could allow operators to move beyond generic campaigns and static plans, towards dynamic, personalised interactions that build loyalty and drive value.

Here are some scenarios where triPica sees the potential where agentic AI could reshape how operators engage with customers:  

• Hyper-personalisation at scale

AI agents can analyse individual usage patterns to deliver timely, relevant offers. Instead of monthly promotions sent to broad segments, each interaction becomes an opportunity for tailored engagement that’s completely relevant to the individual.

• Proactive issue resolution

Predictive models can detect early signs of dissatisfaction and trigger interventions before frustration escalates. This transforms customer service from a cost centre into a driver of loyalty.

• Real-time self-care and assisted support

AI-powered chatbots have potential to resolve routine queries instantly, while CRM co-pilots assist human agents with next-best-action suggestions. The result is faster resolution, lower service costs, and improved satisfaction.

• Service bundling and lifestyle integration

Agentic AI could identify customer behaviours which could be signs that they are likely to adopt third-party services - such as streaming, gaming, or insurance - and when. Offers feel timely and intuitive, not intrusive.

‍

No magic wand for legacy infrastructure

Adopting Agentic AI isn’t merely a matter of deploying new algorithms, and it certainly isn’t a get-out-of-jail-free card for broken or out-of-date BSS.  

It requires a modern foundation: cloud-native, API-first systems that are capable of processing events in real time and integrating seamlessly with third-party services.  

Operators must assess whether their current business support systems (BSS) can support agentic capabilities.  

Most legacy architectures, built for batch processing and siloed operations, can’t be fixed by AI: without the right infrastructure, even the most sophisticated models will be constrained.

When BSS platform is weighed down by monolithic legacy stacks, the business logic is trapped in hardcoded silos. Modern BSS systems are built on microservice architecture, therefore designed for agility. At triPica we use Business Process Management (BPM) engine that enables a modular, building-block approach - perfect for rapid innovation and easy integration of agentic AI.

The BPM engine allows agentic AI to be triggered at any point in the customer journey, triggered by customer or by any agent action - whether it’s initiating a billing process, applying a bonus, modifying a contract, or responding to a customer request in real time. Every interaction becomes a potential trigger for an AI-driven decision, turning your BSS into an intelligent, proactive system rather than a reactive one.

‍

A strategic choice

A transformed customer experience is the prerequisite for growth - and agentic AI offers a credible path forward.  

It’s not as a futuristic vision, but as a practical solution that’s available today.  

The opportunity is clear: modernise the foundation, embed intelligence, and turn CX into the engine of growth.

Learn more about how this transformation is being realised in practice. Contact us.