Description
Summary AI Voice Receptionist for a Multi-Location Clinic (Swedish-speaking) Overview We're looking for a developer to build an AI voice agent that answers inbound phone calls for a clinic chain with three locations. All calls come in through one central phone line. The AI should answer, understand the caller's request, book or reschedule appointments, and route the case to the correct location based on the caller's needs. What the AI needs to do Answer inbound calls in Swedish with a natural, professional voice Understand the caller's intent (new booking, reschedule, cancellation, or general question) Determine which of the three locations fits the caller's situation Create and modify bookings directly in the clinic's existing booking system Answer common questions from a knowledge base (treatments, prices, opening hours) Escalate anything it can't handle to staff, with a summary of the call Send a summary of each handled call to the clinic (e.g. via email or a logging system) Technical context The booking system is Cliento (Swedish platform). They provide an API with developer documentation. A critical question to resolve early: does the Cliento API allow creating and modifying bookings (write access), or read-only? The whole solution depends on this. The system has separate calendars per location — the AI must book into the correct one. Telephony can be handled via Twilio or similar. Voice and conversation logic can be built on a platform like ElevenLabs, or another stack you can recommend and justify. What we need from you An assessment of whether this is technically feasible, especially the Cliento integration Which parts you consider solid vs. uncertain Estimated development time Ideally a breakdown into phases (e.g. base voice agent → telephony → Cliento integration → escalation/reporting → testing) Important We want to start with a pilot/test version before a full rollout, so a solution that can be tested in a limited form first is strongly preferred. Swedish language capability in the voice agent is essential. Please share examples of any voice agents or similar conversational AI you've built previously.