we-home.pe, proptech
WeHome is a property marketplace for Peru where listing is free and contact details stay off the public page.
Service design, architecture and full stack development for our own product, in production at we-home.pe.
Challenge
The brokers who do the work cannot afford to list.
Listing a property on Peru's large portals costs thousands of soles a year. Most of the country's registered brokers do not pay that, so their inventory moves through Facebook groups, WhatsApp chains and signs in the street.
That market has no memory. Messages go unanswered, invented listings sit beside real ones, and the deposit scam is an everyday risk for anyone renting. Nobody holds real closing prices, because nobody records them.
The problem underneath is trust.
In numbers
What the marketplace holds to.
- What a broker pays to list, always
- Free
- Contact details in a public response or the first HTML
- 0
- Locations database behind district and proximity search
- Peru
- In production at we-home.pe
- Live
Approach
We designed the service before we wrote the code.
We started with the whole service: the stakeholders, the journeys people already have, the life of a lead, the trust system and the revenue model. That blueprint became the product's source of truth.
Five principles came out of it, and each one was treated as a design constraint. Verification happens in private, and what a searcher sees is a signal backed by a real process. Conversations happen on WhatsApp, where people already are, and the platform holds the state, the reputation and the metrics.
A broker records stages and closings because it pays them to. Listing is free and stays free; visibility is earned through quality, verification and speed of reply. Every metric answers to a real moment in the service.
What we built
Five pieces, and the service design under them.
- Contact protected by design. No contact detail appears in a public response or in the first HTML. Every contact runs through an authenticated flow that creates the lead first, which keeps the lister's details private and gives the platform sight of the journey.
- Official WhatsApp integration. Leads reach the lister over the WhatsApp Business API, with message templates approved by Meta and the property sent as a PDF.
- AI where it saves work. Listing descriptions and valuation narratives are generated with Claude, so publishing a property takes minutes.
- Geographic search. Peru's full locations database on PostGIS, built for search by district and by proximity.
- Technical SEO. The right HTTP code for every case, 410 for a withdrawn listing among them, and an unavailable-property page that recommends similar homes.
Engineering
Kotlin, hexagonal, and a deploy that rolls itself back.
The backend is Kotlin with a hexagonal architecture, versioned migrations with Flyway and rate limiting on the sensitive endpoints. The frontend is Next.js.
Deployment runs through continuous integration with pre-flight checks, an isolated staging environment and an automatic rollback on any failure. The infrastructure is our own, in Europe, behind Cloudflare. It is the release discipline we use on fintech and compliance work.
Where it goes
Rental first, in the dense districts of Lima.
WeHome grows in stages. Rental comes first, in the dense districts of Lima, where the cycle is short and the trust problem is sharp.
After that come subscriptions for brokers, qualified leads for developers, and a record of what properties actually close at.
Brokers are publishing real inventory at we-home.pe today.
Have a puzzle?
Tell us about the project that standard platforms cannot handle.