Metropolis runs AI-powered computer vision across more than 4,500 parking locations in North America, making it the largest parking operator on the continent. The platform handles checkout-free parking by processing transactions through a recognition layer that identifies vehicles and members in real time - more than 1 million new members monthly, serving over 50 million customers total. The company has raised $1.6 billion and employs more than 22,000 people.
For security and engineering teams, the threat surface is significant: a computer vision system managing physical access and financial transactions at scale, processing biometric-adjacent vehicle recognition data across thousands of sites with 99.9% uptime requirements. The core technical domains are computer vision and machine learning, operating as infrastructure for a physical-world economy rather than a purely digital one. The intelligence layer does more than identify - it enables precognitive, personalized experiences tied to real spaces and real objects.
When a system like this goes down, the failure mode isn't an error page - it's a parking garage full of people who can't leave. The data pipeline handles identity, payment, and movement patterns at a volume that makes it a high-value target. Security roles here sit at the intersection of ML model integrity, IoT sensor networks across thousands of physical locations, payment processing, and PII handling at a scale that touches a meaningful fraction of the North American driving population.
