A Montréal startup is using generative AI to try to speed up one of the most laborious parts of real estate transactions in Québec—and eventually, the rest of the country.
The news: Proptech startup Paraito announced on Monday it had raised $2.65 million CAD in equity pre-seed funding, led by Inovia Capital and with participation from Boreal Ventures. It has also raised roughly $650,000 from friends and family, including individual notaries and lawyers.
The startup says its AI-driven platform can automate title searches. Title searches are a verification check on public information about a property—including outstanding liens or claims against it, which could interfere with a clean sale.
From the source: “When we launched, I think we sent 35 emails to notaries and got 29 positive responses,” Paraito co-founder François Arbour claimed in an interview with BetaKit. “They were like, when can I use this? We saw it was a real pain point.”
Following the thread: In Québec, property transactions require several paperwork-intensive steps, including legal document review and registration with provincial authorities—many of which are outsourced to notaries.
According to Arbour, the title search is the most tedious step in that process. He said that eight or nine months ago, AI models weren’t good enough to perform work like analyzing old handwritten documents, but now, he’s confident the models they use are “ready for prime-time.” Paraito is selling its platform, which for now is only available in French, on a per-search basis to small and large notary firms.
Final thought: According to federal jobs data, notaries have broadly been in short supply in Montréal and Québec, with more openings than workers. Radio-Canada reported earlier this year that an increase in real estate transactions and more retirees has created a months-long backlog of customers waiting to consult with notaries in Trois-Rivières, Que. Paraito is positioning itself to ease some of that pressure by claiming its product can help firms process three times more files without hiring more people.
Feature image courtesy Romain Dancre via Unsplash.