Immobiliare.it builds conversational real estate agent in days using ElevenLabs

Immobiliare.it is the largest real estate platform in Italy. As part of its transformation into an AI-native company, the team explored new ways to improve the experience for both users and agents — starting with how people interact with property listings.

They built a conversational voice agent using ElevenLabs’ Agents Platform to solve a long-standing problem in property marketplaces: sellers want to avoid spam calls, while buyers are often frustrated by a lack of response. Immobiliare’s new voice agent works on behalf of sellers to answer listing-specific questions, gather lead information, and summarize the interaction. It runs 24/7, responds in natural Italian, and was built in just days using ElevenLabs.

From JSON to voice in minutes

The original goal was to create an assistant that could take listing data in JSON format, hold a natural conversation with users, and return a structured summary of the interaction.

While evaluating tools, the team started with OpenAI APIs. But when more flexibility and Italian language support was needed, the assistant itself, running inside a prototype built with Lovable, recommended ElevenLabs.

They set up an ElevenLabs account, created a voice agent, and had it integrated in five minutes. The demo was shared with product leadership and quickly approved for further development.

Building with speed and autonomy

The team followed a two-stage build process:

• Prototype: Using n8n and ElevenLabs, they created a working backend that handled voice input and response generation using live listing data.
• Production: Once the prototype was stable, engineers implemented a scalable backend integration. Handoff was seamless thanks to early prompt tuning and data validation.

This approach enabled rapid iteration with minimal engineering overhead.

“Normally, after you launch a feature, there’s a short moment when the dust settles. You catch your breath, you wait for data to come in, you plan the next iteration. Not this time. With ElevenLabs, our feedback loop went from weeks to hours… Thanks to the transcripts, we could take a real conversation, tweak the prompt in staging, test multiple variations, and then push the change live to production. It felt like fine chiseling: every small adjustment brought the agent closer to behaving the way we needed it to.” – Gaia Businaro, AI Product Manager

Better experiences for users and agents

The AI agent helps Immobiliare.it solve two critical problems:

1. User response quality: Every user receives an immediate, accurate, and human-sounding reply to their listing queries.
2. Lead qualification: The agent asks questions to determine how interested the user is - filtering for serious inquiries.

Adoption numbers confirm this shift, with more than 70% of sellers opting in. Lead qualification has risen from 19% to 63%, and the share of users providing a phone number for follow up has increased from 42% to 73%. Roughly 80% of users also rated the feature positively and early feedback indicates that buyers are more comfortable sharing personal details over voice than text, generating stronger and more actionable leads.

“These early outcomes confirm that the ElevenLabs-powered voice assistant is the right solution to the problem we set out to solve. They give us the confidence to rapidly scale the approach across more listings and additional use cases.” - Paolo Sabatinelli, Chief Product Officer

The result is higher satisfaction on both ends: users get the information they need quickly, and agents receive better leads with clear intent data.