Product and Services Roundup: Bien'Ici, Jitty, Propic

This week's Product roundup starts in France...

Europe: Bien'Ici adds Display Local feature to improve visibility for agents

French portal Bien'Ici has launched "Display Local", a new feature that boosts visibility for realtors advertising on its site.

Subscribers will receive a premium placement at the top of search results. More importantly, there are only two spots available per municipality.

Display Local adds branded interchangeable banners according to the realtor's needs. Options include Valuations, Rental Management, Luxury Properties, Seasonal Rentals and Recruitment. However, realtors can only change their banner once per year.

Didier Monnet, CEO at Bien'Ici, said:

"Thanks to the Local Display offer, you can position your real estate agency as a local reference by promoting your expertise to sellers and lessors in your catchment area. To guarantee maximum visibility and limit competition, this advertising system is only open to two agencies per zip code.

"This advertising system allows real estate professionals to promote their brand and services via a thematic banner of their choice with their logo and personalized with the agency’s colors, and to benefit from a Premium placement in the #1 block in search results.

"By subscribing to this offer, the real estate agency is guaranteed to be visible to half of the Internet users who are looking for a property in its area or who are planning to sell."

Australia: View Media Group launches conversational AI for listings

View Media Group's Propic has released new capabilities for its AI platform 'Claire', introducing a voice-enabled interface and multilingual functionality that allow users to speak directly with property listings—without needing an agent present. The update reflects a strategic push toward voice as a core user interface for property discovery and positions Propic as a frontrunner in automating real estate communication at scale.

The upgraded Claire enables property seekers to tap a speaker icon on a listing and initiate a live, natural language conversation about the home and surrounding area. The assistant responds with contextual answers drawn from listing data, CRM systems, and local market knowledge. The platform now supports 58 languages, with instant translation back to English for agents on the receiving end—effectively removing language as a friction point in buyer engagement.

Propic CEO Jeffrey Gray said:

"We predicted 2025 would be the year voice went mainstream. With consumers used to interacting with assistants like Alexa and Siri, it was time they could simply talk to a property 24/7. And in a multicultural market like Australia, expecting everyone to communicate in English just isn’t good enough anymore. We’ve built Claire to speak every buyer’s language—literally."

Claire’s voice interface is underpinned by compliance safeguards and algorithmic controls to ensure accurate and brand-safe responses. The assistant also integrates directly with CRM platforms, enabling it to retrieve real-time data and documents during conversations.

Propic, founded by Gray in 2019, has established itself as a category leader in real estate-focused generative AI. Claire currently supports major brands including McGrath, Nelson Alexander, Place, and Highlands. Over the past 12 months, the platform has handled more than $13 billion in property transactions, automating tasks ranging from enquiry triage and buyer nurturing to content generation and lead qualification.

The latest updates are part of a broader product roadmap aimed at embedding voice deeper into the user journey. Propic is positioning Claire as a multilingual digital agent capable of bridging not only time and scale gaps, but also cultural and linguistic divides.

United Kingdom: Jitty upgrades search function for its own limitations

Challenger home search platform Jitty has upgraded its natural language search to admit defeat when it can't give accurate answers.

According to an email from Jitty:

We’ve made a big upgrade to how our natural language search works. It’s quicker, more accurate and better at understanding what you're actually asking...and when it can't answer properly.

Before, if you searched for something like low crime rates, the model would treat it like a visual search. That works for “pink bathroom.” Not for “good schools.”

Now, if we can’t support a request, it just says so. That means clearer results for you and, more importantly, better signals for us. We can see where the gaps are more clearly, which helps us prioritise the right features, faster. Shorter feedback loops mean we can build a better product for you.

We also rebuilt how we prompt the model. Prompts are now generated based on the filters and values we actually support, so answers should feel a lot more useful than before.