Integrating AI to create a more proactive user experience

Integrating AI for a Proactive User Experience

Belgium’s leading telecom and digital services provider, serving millions of customers. In the telco industry, staying competitive isn’t just about network quality; it’s about the customer experience. Proximus saw an opportunity to leap ahead by leveraging AI to create smarter, more proactive user interactions.

Over the course of several weeks, we co-created and prototyped six high-impact use cases that demonstrated the power of proactive AI. Each use case was designed to enhance user experience and operational efficiency. All of these experiences were built on Nimble’s modular agentic infrastructure, allowing us to rapidly spin up specialized AI agents that work together to deliver intelligent, cohesive outcomes.

Opportunity

A clear need

Even in a high-tech sector, many customer experiences remained frustratingly reactive. Users would only get support after a problem hit, a surprise billing shock or a service outage. Internally, Proximus teams felt this too: support centers swelled with repetitive inquiries, and opportunities to delight or upsell customers were missed. The pain points were clear: customers wanted help before they knew they needed it, and Proximus wanted to provide it.

AI as an opportunity

Generative AI promised a way to anticipate and act on customer needs in real-time. But doing AI right is hard,  nearly 90% of AI projects fail to deliver value. Proximus wasn’t about to become a part of that 90%. Instead, they started an ambitious AI Discovery Track to crack the code on proactive customer experience. The mission: identify the highest-impact use cases and prove them out fast, before interest could fade or the window of opportunity closed.
solutions

A fast paced AI Discovery Track

Proactive support
In-home experience
AI-enhanced in-store support
Rights delegation
Self-service upgrades
Insights dashboard
POC 1

Proactive support

An AI support agent that predicts common issues and addresses them before customers reach out. Think alerts about “bill shock” (unexpected high charges) with one-tap solutions, or a heads-up on network disruptions with automatic troubleshooting. This turns customer support from a reactive hotline into a preventive concierge service.

POC 2

In-home experience

A smart “homefeed” agent that personalizes the in-home digital experience. It learns from your usage (TV, internet, smart devices) and proactively suggests content or adjustments. For example, it might pre-load the evening’s football match in HD knowing you’re a fan, or remind you to set up a Wi-Fi guest network before your friends arrive. It’s like a digital butler making your home life smoother.

POC 3

AI-enhanced in-store support

Bringing AI intelligence into Proximus shops. We prototyped a co-pilot for store staff that instantly pulls up relevant info and recommendations as they talk to a customer. Imagine a sales agent who, with AI help, already knows you called about an issue last week and preps a tailored solution before you even ask. The result? Faster, hyper-personalized service in-store, with AI working hand-in-hand with human experts.

POC 4

Rights delegation

Simplifying how customers manage who can access their services. This use case introduced an AI-driven flow to delegate account rights securely to family or colleagues. For example, a parent could grant a teenager limited control over the home Wi-Fi or a business owner could allow an IT assistant to manage the company’s mobile plans, all through a natural-language chat with an AI that ensures proper permissions are set without hassle.

POC 5

Self-service upgrades

An intelligent self-service agent that guides customers through solving issues or changing services on their own. Rather than combing FAQs, users get a conversational AI that already knows their context. “Why is my internet slow?” might prompt the agent to run a quick diagnostic and walk the user through a fix, or offer to upgrade their plan if needed. It’s DIY support, supercharged with AI, reducing calls to the helpdesk and empowering users.

POC 6

Insights dashboard

A proactive insights engine for both customers and internal teams. For customers, it visualizes their usage and spots ways to save or get more value (e.g. “You haven’t used your international minutes – how about a smaller plan that saves €10/month?”). For Proximus managers, it aggregates live data from all these AI interactions into a dashboard of trends, pain points, and opportunities. It’s not just an analytics tool; it’s an AI that flags what matters most, in real time.

Proactive support
In-home experience
AI-enhanced in-store support
Rights delegation
Self-service upgrades
Insights dashboard

Technical Foundations

How do you deliver six AI-powered prototypes in a matter of weeks? Technical underpinnings were key: Nimble brought a ready-made stack and a few AI tricks up our sleeve.

Architecture overview
GenAI Pilot Environment:  Cloud-native setup integrating LLMs with Proximus’s data in a secure environment.
Microservices & AI Agents: Each use case deployed as a set of modular services and agents.
Central Orchestration Layer: Smart dispatcher coordinating tool-calling and intent triage.
Tools & technologies
LLMs: full model independence (use-case specific)
Frontend: Modern frameworks for mobile and web dashboards (Mendix-compatible webviews & react)
AI Copilots: Vibecoding tools, mainly focused on speeding up the coding work on delivering frontend code (summarization, ideation, coding, and presentation building)
ai-in-the-loop
GPT-based tools summarized interview transcripts.
Workshop ideas distilled and structured by AI.
Coding accelerated by generative AI assistants.
Prototypes co-built by PM’s and AI tools (e.g. one prototype in just 45 minutes).
Presentation materials (slides, narratives) partially auto-generated.
outcomes

Delivering real value

For customers

Customers can expect a smarter, more proactive service experience from Proximus. With six working AI prototypes developed at record speed, the groundwork has been laid for solutions that anticipate customer needs before they arise, whether you’re talking to a service rep or navigating digital touch points. This means faster support, more relevant communication, and an overall smoother experience powered by intelligent automation.

For Proximus

For Proximus, this track set a new benchmark for AI speed and agility. Their team didn’t just build prototypes, they validated business value, secured stakeholder buy-in, and aligned their roadmap around real, impactful use cases. Internally, teams were empowered by hands-on exposure to cutting-edge AI tools, unlocking momentum to keep building. As a result, Proximus has positioned itself as a first mover in the proactive telco space, ready to lead with innovation.
sneak peak

What’s next?

The prototypes are just the beginning. The next step is turning these six promising AI use cases into fully integrated products, bringing them from lab to living room. With user testing, ongoing R&D, and a new internal AI task force, Proximus is laying the foundation for an AI-powered future; one where autonomous agents proactively solve problems, optimize plans, and unlock new value for customers, often before they even ask.
This AI Discovery Track marks the start of a bigger shift: from reactive services to agentic experiences: where digital systems act on behalf of users, not just respond. As Nimble continues to guide this transformation, every step forward brings Proximus closer to a radically smarter, smoother, and more intuitive customer journey.
The future is unfolding, and it’s being built one proactive interaction at a time.
Features

The right things,

built right.

Connectivity made hassle-free, ensuring smoother operations and better customer service.

Multi-agent setup
Hyper-personalised
notifications
Anticipatory homefeed