Very AI

About Very AI

Very AI is a technology company founded in 2025 by Zach Meltzer, developing infrastructure to combat digital fraud and synthetic media. The company is headquartered in Los Angeles, California. In March 2026, Very AI announced a $10 million seed funding round led by Polychain Capital, with participation from the Berggruen Institute, Anagram, and an angel investment from Solana co-founder Anatoly Yakovenko.

The founder and CEO, Zach Meltzer, has a background in building identity and credentialing systems for the crypto industry, having previously helped scale the crypto loyalty platform Galxe to millions of users. This experience exposed the limitations of existing verification methods against the backdrop of increasingly sophisticated AI-driven fraud, which motivated the creation of Very AI. The company's Chief Science Officer, Hua Yang, is a researcher in palm biometrics with numerous academic publications and patents in the field.

Very AI operates on a B2B model, providing its services to businesses like fintech companies, crypto exchanges, government institutions, and social platforms. Revenue is generated based on monthly user verification volumes. The core of Very AI's offering is a "Proof of Reality" platform that uses a smartphone's camera for identity verification through palm biometrics and liveness detection, requiring no specialized hardware. This system is designed to provide a more secure alternative to traditional methods like facial recognition, as palm prints are less publicly exposed and harder to forge. The technology converts palm scans into irreversible mathematical patterns stored on the Solana blockchain, leveraging zero-knowledge proofs (ZKP) to ensure user privacy during verification. The company reports a false acceptance rate of approximately 1 in 10 million for a single hand, which drops to 1 in 100 trillion when both hands are used.

Beyond human verification, the platform includes tools for deepfake detection and authenticating digital content to prove it was created by a verified human. To further its research, Very AI has partnered with Northwestern University's Human-AI Collaboration Lab to study deepfake detection.