Service-sector SME (B2C and B2B mix)

A Swiss-German Voice AI Receptionist for a Service-Sector SME

Inbound calls answered in Schweizerdeutsch, leads qualified, and appointments booked straight into the calendar — around the clock, without adding headcount.

By Nico Jaroszewski · Founder, AETHER DigitalPublished on 2 May 2026
Industry
Service-sector SME (B2C and B2B mix)
Size
small operations team
Region
German-speaking Switzerland
A Swiss-German Voice AI Receptionist for a Service-Sector SME

The challenge

The client was losing identifiable revenue every week because calls went unanswered outside office hours and overflowed during peaks. Adding a permanent receptionist was hard to justify for the load, and an outsourced call centre could not handle the dialect or the bookings system. The team had tried generic chatbots on the website with poor results — most prospective customers in their segment still pick up the phone.

The brief was specific: answer the phone in natural Schweizerdeutsch, ask the right qualifying questions, write the appointment into the existing calendar with the correct service type and duration, and hand off cleanly to a human when the situation called for it. Never make a customer feel they had been routed to a robot.

Our approach

We scoped and built the agent through our AI voice and chat agent practice, with integration delivered alongside business process automation so the agent reached into the systems the team actually used. The conversation flow was designed by listening to recorded historical calls and writing the qualification logic the receptionist would have run themselves on a good day.

Voice work centred on Schweizerdeutsch comprehension and a natural-sounding response voice that the team's own customers tested in a closed pilot. The agent runs against a constrained policy: it can ask, confirm, book, reschedule and cancel, but it does not improvise outside its mandate. When a caller's request falls outside policy, or when the agent's confidence drops, the call is warm-transferred to a human with a written summary already on screen.

The booking integration writes directly to the client's calendar with service-specific durations and buffers, and a confirmation SMS goes out automatically. Every call is transcribed and tagged so the team can review quality without re-listening.

Outcome

Within weeks the client was answering effectively every inbound call, including outside business hours, with the great majority handled end-to-end by the agent. Booked appointments coming through the agent showed the same show-rate as those booked by the human team, which the client treated as the real measure that the qualification was working.

The team's role shifted from picking up the phone to handling the calls that genuinely need a human. The Swiss-German voice has become part of the brand experience rather than a workaround.

Stack used

Realtime voice LLMSchweizerdeutsch ASR and TTSConversation policy engineCalendar integrationSMS confirmationCall transcription and taggingWarm-transfer handoff

Duration

approximately 10 weeks

Related services

Discuss a similar challenge with us

Contact Us