Profile
The Face at the Frontier
You are already inside his world. The passport gate that waved you through in two seconds - that was probably Paravision. The border agent who barely glanced at your face before nodding - same. Doug Aley runs the company that the government agencies and security firms call when they need the numbers to actually work.
Paravision holds the #1 global ranking on NIST's Face Recognition Vendor Test - the most rigorous independent benchmark in the industry. They're the only American company that has consistently cracked the global top 10. Their error rate for one-to-many face identification dropped by 90% in four years. These are not marketing figures. NIST tests everyone, publishes everything, and doesn't take calls from PR departments.
Aley came up through the long route. He co-founded his first company at 19 - a web accessibility firm called SSB BART Group (now Level Access) before ADA compliance was anything more than a nuisance footnote for most tech teams. Then came stints at Amazon, a string of startups, a run as VP of Product at Zulily during the years the company went from $100 million to $700 million in revenue ahead of its IPO, and a brief, high-profile turn as CEO of Shuddle - the "Uber for kids" - in 2015.
He arrived at the company that would become Paravision in 2016 as an operating partner at Atomic Ventures - the venture studio where the original entity was incubated. He took the CEO role in 2018. The company was still called Ever AI then, running a consumer photo app. The facial recognition models built in-house were almost incidentally good. Enterprises started asking to license them. The consumer app got shut down. The company pivoted, rebranded, and went after the global identity infrastructure market.
What followed was a methodical, measurable climb up the NIST leaderboard - not through PR, but through engineering. Paravision submitted to the vendor tests repeatedly, published results transparently, and built products that enterprise customers and government agencies could actually deploy. By 2022, they'd claimed the #1 spot on the 1:N identification test - the hardest challenge in the category, matching a face against a large database rather than just verifying a 1:1 match.
"We regularly turn down business when we are unconvinced of the responsible use of this technology. So while countries like China are leading the way in the implementation of face recognition, it is unlikely that we will ever operate there."
That quote gets cited a lot - because it's the kind of thing tech CEOs almost never say out loud, let alone mean. China is the biggest deployment market for face recognition technology in the world by orders of magnitude. Aley walked away from it. The company's ethics policy isn't a landing page - it's a deal filter that's cost them revenue.
The company's current trajectory runs right through the problems that matter most in 2025: liveness detection (is this a real face or a photo?), deepfake detection (is this face real or AI-generated?), and cross-domain identity convergence (can one biometric layer work across physical security, digital login, and government ID?). In March 2025, Paravision launched Liveness 2.0 with passive detection - no active challenge, no instruction to blink or turn. In June 2025, Deepfake Detection 2.0 arrived, targeting the wave of synthetic faces proliferating through identity fraud pipelines.
Aley's own description of his role at the early company: "I would say all of us early folks at Paravision kind of feel like founders, but not technically founders." That framing - the deliberate extension of founder identity to the full early team - says something about how he builds. Not by title, not by equity cap table, but by culture of ownership.
His LinkedIn bio lists his priorities as: "Husband, Father, CEO." In that order. He lives in Greenbrae, Marin County, with his wife Susan and their two sons. It's the kind of biographical detail that would read as a press-trained softening if it weren't for the consistent pattern across his career: he builds things that work, stays until the work is done, and moves with intention.
