Alquilia Case Study, an AI CRM That Qualifies Tenants over WhatsApp

Project type
AI CRM for rental agencies
Years
2026
Deliverables
Website and dashboard, WhatsApp assistant, document reading, and plans and billing
The Alquilia leads panel in a browser and a tenant's WhatsApp qualification on a phone

Alquilia is a program for rental agencies that answers anyone who asks about a flat over WhatsApp. It picks up the people who come from the property portals, asks them the agency's questions, asks for their documents and gives them a score, so the agency only spends its time on those who qualify. We built all of it for Alquilia, from the website to the assistant.

A phone in an empty flat for rent with a candidate's WhatsApp conversation in Alquilia's dashboard.

From the portal to WhatsApp

When someone asks about a flat on Idealista, Fotocasa or any of the ten portals it recognises, Alquilia reads the notice that reaches the agency's email and writes to them over WhatsApp from the agency's own number, at night and at weekends too.

Only those who qualify

It asks the questions the agency chooses, understands their voice notes, asks for the documents their job calls for and reads them. With that it gives them a score from 0 to 100 against the agency's criteria, and the agency gets a notice when the candidate is ready.

A phone on a street in Madrid with the questions the agency chooses for its candidates.
A phone in a café with a test conversation with the assistant.
A phone in a flat's living room with the flat's page in Alquilia, its floor plan and its price.
A phone on a terrace with a candidate's score and the checks behind it.

All in one dashboard

The agency sees its candidates, its flats and every conversation in one dashboard, and can write from there. Before switching the assistant on it tries it in a test conversation, and each agency has its own team and plan.

A laptop in an agency's meeting room with Alquilia's analytics for the last 30 days.

How it's built

  • Website and dashboard

    React and TypeScript, for the public website and each agency's dashboard.

  • Server

    Java with Spring Boot and PostgreSQL, with each agency kept apart from the others.

  • WhatsApp

    The agency's own WhatsApp, connected with a QR code through Evolution API.

  • AI

    Llama models on Groq to talk and read documents, and Whisper for voice notes.

  • Email

    It reads the portals' emails in the agency's Gmail or Outlook and sends its notices with Resend.

  • Documents and billing

    Documents and photos stored in Amazon S3, and plans and invoices with Stripe.

In numbers

portals it picks up candidates from
10
types of document it recognises
34
checks behind every score
6
Next case: Boathous