Deck raises $12M to ‘Plaid-ify” any website using AI

Summary

Deck, a startup that claims to be building “the Plaid for the rest of the internet,” has raised $12 million in a Series A funding round — about nine months after closing its seed financing, it tells TechCrunch exclusively. The new raise, led by Infinity Ventures, brings Montreal-based Deck’s total raised since its January 2024 inception to $16.5 million. Golden Ventures and Better Tomorrow Ventures co-led its seed raise. Deck claims that it is building the infrastructure for user-permissioned data access — across the entire internet. Its browser-based data agents “unlock” the data from any website through automation. To put it more simply, Deck helps users connect any account online and aims to turn the information into structured, usable data, with full user permission. President Frederick Lavoie, CEO Yves-Gabriel Leboeuf, and CTO Bruno Lambert (pictured above, left to right) co-founded Deck in June 2024. The startup’s approach is to treat the web itself as an open platform. It operates under the premise that users have “tons of valuable data” locked behind usernames, passwords, and session-based portals with no real way to share it securely. Deck hopes to change that. “Just like Plaid gave developers an easy, secure way to access bank account data with user permission, Deck does the same for the 95% of platforms that don’t offer APIs such as utility portals, e-commerce backends, payroll systems and government services,” Leboeuf told TechCrunch. Its goal is to make it easier for developers to access the data users already have without all the manual work When a user connects an account, Deck’s infrastructure handles everything behind the scenes. Its AI agents log in, navigate, and extract the data “just like a human would — but faster, more reliably, and at scale,” said Leboeuf. It then generates scripts to keep those connections live and reusable without AI involvement going forward. “Companies use Deck to eliminate the friction of getting their user data from pla...