Problem

In Switzerland, rented apartments come with fitted kitchens, resulting in 70% higher property management inquiry volumes compared to Germany due to appliance repairs. Immofix manages hundreds of thousands of these inquiries for white-label clients. The volume was crushing, but the client was skeptical of a long discovery phase, citing that their processes were already "mapped out" despite relying on inconsistent human inputs.

Solution

We bypassed the traditional discovery phase to respect the client's urgency and budget. We proposed and built a rapid "Data Gathering" MVP. Instead of a full end-to-end automation immediately, we deployed a low-risk AI Voice Bot dedicated solely to intake.

The AI answers calls, transcribes the complaint, and uses vector embeddings to classify the incident (e.g., "Dishwasher - Pump Failure" vs. "General Plumbing"). This provided immediate relief to the hotline while gathering the structured data necessary to train the full automation pipeline later.