Les bots sont là, oh là là !

Vicky, Freddy, Cycly, and Lisa

They don’t wear name tags and don’t need helmets. Yet Vicky, Freddy, Cycly, and Lisa are among the best members of their customer service teams. Vicky answers questions at the Aargau cantonal vehicle registration office, Freddy provides support to the Südostbahn customer service team, Cycly shares her knowledge about recycling, and Lisa directs Mobilezone customers to the right representative. They are increasingly supported by AI systems that do more than just answer questions—they trigger the next steps or even carry them out directly. The twelve speakers with hands-on experience clearly demonstrated how rapidly the scope of these “AI” colleagues is expanding. They transcribe and file conversations, verify documents, open accounts, resolve IT issues, draft emails, and provide employees with the appropriate context during conversations. The crucial question, therefore, has long since ceased to be whether and when AI will replace customer service representatives. What is more exciting is seeing how companies orchestrate people, processes, knowledge, and technology so that customers truly feel a difference. Concrete examples from the public sector, the gambling industry, insurance, retail, mobility, media, and recycling have provided remarkably concrete answers to this question.

The Digital Connection to the Aargau Vehicle Registration Office

Registering, changing, or deregistering a vehicle: about half of the phone inquiries received by the Aargau Vehicle Registration Office are recurring. This is exactly where Vicky comes in. Initially, the voice bot could handle only five questions; today, it can handle 23 types of requests and has been systematically trained based on customer feedback over the course of seven optimization cycles—including for dialects and non-native German. Customer adoption is growing: one-third of customers already use Vicky outside of business hours—or choose her over waiting in line. The next step will be to integrate additional processes, channels, and administrative services. The goal is to establish an interconnected AI platform capable of keeping pace with the strong growth of the vehicle fleet without a corresponding increase in staffing requirements, explained Sandro Fanti, head of specialized services and transformation at the Aargau Motor Vehicle Department.

“Service Makes the Difference” – Swiss Casinos Automates Processes Where It Enhances Customer Convenience

Swiss Casinos operates four land-based casinos and one online casino. Competitive pressure is particularly intense, especially in the online sector. Raphael Grossenbacher, head of customer service at Swiss Casinos, responds: “Quality and the product offering get you in the game. It’s the service that makes the difference.” He heads a department available 24 hours a day in four languages. The context is complex: this regulated industry requires rigorous controls regarding identity verification, anti-money laundering, and player protection. Its virtual agents now handle more than half of the volume. In the chat service, AI handles as much as 65% of the volume. This frees up staff time to handle exceptional cases, complex situations, and relationships with regular customers. Grossenbacher’s guiding principle is as simple as it is relevant: automate where it adds convenience—and rely on humans where judgment and flexibility are needed.

AXA Turns Conversations into a Source of Learning

At AXA Switzerland, the value of AI starts with accurate transcription. Gennaro Chirico, Product Owner for Conversational AI, explained why general speech recognition isn’t sufficient for this purpose. Product names and technical terms require targeted training—especially when “AXA” suddenly turns into “Roxanne” in the transcript. This fine-tuning of AXA’s vocabulary is paying off. High-quality transcriptions and automatic conversation summaries save one to two minutes per call; in total, this represents a reduction in workload equivalent to about two full-time positions. At the same time, the insights gained are enhanced: the data supports quality management, training, fraud prevention, and the analysis of recurring customer requests. “In everything we do, the customer experience is at the heart of our efforts. AI helps us transform the conversations we have into conversations from which we learn.”

Migros is rethinking IT support by starting with the customer’s needs

Sven Bühlmann, head of the Group IT Service Desk at the Federation of Migros Cooperatives, articulated the central question behind his project’s design as follows: How can we simplify IT support when, behind the scenes, everything is becoming more complex? In the future, users will no longer need to know the specific system, the correct form, or the underlying process. They describe their request verbally: the platform identifies the intent and triggers the appropriate action. The AI agent is capable of detecting problems and carrying out steps, while employees can, if necessary, validate or take over. “This is important because our user base is set to double, from 30,000 to 60,000.” The vision is “Invisible IT”: the problem is resolved before it even affects the user.

SOB is moving toward the next level of service

Sonja Zuberbühler, head of customer service at Schweizerische Südostbahn, discussed the transition from a traditional management tool to a modern service platform. Freddy, the virtual colleague, is tasked with improving service quality, easing the daily workload of employees, and making knowledge more readily accessible. An unflinching look back at the journey so far proved particularly valuable: new technologies do not deliver their full benefits immediately upon implementation. A platform capable of learning, realistic expectations, and the willingness to continue developing processes after launch are decisive factors.

Service Transformation at Assura

Innovation and customer experience are also central to Assura’s priorities. The health insurance provider migrated its contact center to the cloud to fully leverage the potential for customer experience (CX) and employee experience (EX). Today, conversations are automatically transcribed, summarized, and transferred directly into the CRM. Yassir Madhour, CIO of Assura, estimated the time saved on documentation at about one minute per conversation—which, with nearly one million calls per year, represents a significant gain. This example also highlighted the importance of governance in a regulated environment. Sensitive medical data is removed from the call system or the transcript; compliance and risk management were built in from the start. Once the foundations are firmly in place, the next steps include AI-powered search, predictive scheduling, chatbots for simple processes, and voicebots to replace traditional IVR menus. Madhour’s motto: get the fundamentals right while continuing to innovate.

From KPIs to “Why”: How Callpoint Evolved from an Outsourcing Provider to a Trusted Advisor

Four years ago, when Callpoint analyzed its IT environment, the company identified 130 pain points: media disconnects, scattered knowledge, manual tasks, and numerous siloed solutions. For José Manuel Taboas, Chief Operating Officer (COO) of Callpoint AG, the move to the cloud was not a decision driven by telephony, but one focused on customer experience (CX). Phone, email, chat, routing, workforce management, reporting, and AI all needed to be consolidated onto a single platform. The results are compelling. For complex projects, training time has dropped from 130 to 75 hours, and time to go live has decreased from twelve to eight weeks. Average handling time decreased by 12.5%, while quality increased by 10%. “AI must provide insights to humans. And the contact center must shift from key performance indicators (KPIs) to the ‘why’: it’s not just about seeing what happened, but understanding why” Mr. Taboas cautioned, however, against hasty automation: a bad process, when combined with AI, simply becomes a bad process that runs faster. Simplification and standardization come before automation. Only then can Agentic AI execute tasks and see them through to completion.

Ringier combines three AI stacks

Gregor Ottiger, head of the customer service center and back-office services at Ringier, opened the conference with a delightful anecdote: one of the first voice-based AI systems confused the Swiss “Abi” with the German “Abitur.” The lesson to be learned remains: linguistic reality trumps any demonstration. Systems must be tested with dialects, technical language, and complex everyday situations. Today, 80% of cases are already partially or fully automated, saving valuable time that can be devoted to advising clients. However, voice AI, email AI, and case summarization were developed on three separate platforms. Each solution has delivered results—but has also created new interfaces, new metrics, and new responsibilities. Today, Ringier is consolidating these systems where the knowledge, processes, and data reside. This is because AI is only effective within its specific context, and scaling it up requires a central hub for management and measurement.

Swiss Post Saves Time to Listen

Madelon Bakx, Product Owner for conversational services at Swiss Post, highlighted an often-underestimated workload: agents must simultaneously listen, type, search, and document. To enable employees to develop their “superpower”—empathy—Swiss Post implemented a voice AI solution in less than two months. An initial use case makes it easier to track shipments in Postino and reduces redundant administrative tasks. The true benefit, however, cannot be measured solely in terms of automation rates. “When the attention lost between the conversation and the systems is reduced, there is more room for genuine listening and empathy. In this case, AI doesn’t create distance from customers, but rather provides better conditions for building rapport,” explains Mr. Bakx, drawing on his experience.

Helsana: A Human Approach Meets Scalability Thanks to AI

The peak season for health insurance providers begins next week. Patrick Angehrn, Product Owner for Realtime Interaction Management, and Tanja Bosshard, head of the Training team, demonstrated how to evaluate quality in a more nuanced way in this high-volume sector. Transcripts form the basis for AI-assisted analyses, evaluations, and feedback. Coaches remain the point of reference. During the pilot project, human expert judgment and AI evaluation were compared iteratively. The prompts, model, and transcripts were continuously calibrated. Helsana thus combines two strengths: AI evaluates consistently, adaptively, and transparently according to jointly defined standards. Humans contribute their experience, empathy, and judgment. The technology therefore expands the sample size; responsibility for quality remains in human hands. “Humans set the standard. AI scales it up,” explains this dynamic duo.

Recycling Know-How for Everyone

With a team of ten people, Swiss Recycle has been compiling recycling knowledge for the general public, municipalities, businesses, collection points, and schools for about 30 years. Viviane Pfister, co-CEO, demonstrated how the Cycly bot makes this knowledge accessible 24 hours a day. The chatbot answers questions on the website; for email inquiries, the AI drafts responses that employees then review, adapt, and send. Analyzing these inquiries not only helps identify the materials and target groups that generate the most questions but also provides insights into areas for improvement regarding the website and outreach efforts. The main lesson learned may seem trivial, but it is crucial: “Content management is a long-term effort. A knowledge base must remain up to date, and 100% accuracy is impossible without constant monitoring,” explains Ms. Pfister.

Mobilezone Turns Voice into an Interface

From clicking to typing, then to speaking: AI is becoming the new user interface. Mobilezone does not view voice AI as just another channel, but as a logical evolution beyond the era of websites and apps. “After clicking and typing, speech is once again becoming the central interface,” the two department heads assert with conviction. For Lukas Degrandi, Director of Customer Service and Quality, and Fabian Kistler, AI and Automation Engineer at Mobilezone, the voice agent is named Lisa. She identifies the purpose of the call and summarizes conversations in the blink of an eye. The key to maintaining stable customer satisfaction levels? “There are no more wait times, and human agents take over with a better understanding of the context. ” The next steps range from identification and authentication to autonomous handling of requests. In the long term, Lisa should also provide support in the areas of consulting, after-sales service, and upselling and cross-selling thanks to real-time information.

These new colleagues need guidance

What have we learned after twelve practical examples and 15 presentations at Speakers Corner? Bots and AI agents are becoming indispensable parts of hybrid teams. But even virtual employees need to be properly integrated: reliable knowledge, clear authorizations, defined escalation procedures, quality control, and ongoing training. The more autonomously they operate, the more important data quality, process design, and governance become. That doesn’t mean humans are disappearing from customer service. Their role is evolving. Machines can handle routine tasks and documentation; context, relationships, empathy, and judgment remain fundamental human skills. This is precisely where the day’s shared conclusion lies: Good AI doesn’t relieve humans of the need to think. It gives them more space to do so. And what about the new team members? Vicky, Freddy, Cycly, and Lisa should soon have some company. What will matter most is not how “human” their names sound, but how companies manage to harness their capabilities—and whether customers and employees view this development positively.

Event Overview

We look forward to seeing you again and sharing even more best practices! The best opportunities to do so: On November 19, 2026, the Swiss Customer Relations Awards will be presented and the next Swiss Customer Service Summit will take place on September 23, 2027, once again at the Kongresshaus in Zurich. Stay tuned…