Oasis Labs

Oasis Labs builds privacy-preserving data and artificial intelligence technology based on a combination of cryptographic and statistical techniques, including differential privacy, secure multi-party computation, homomorphic encryption and zero-knowledge proofs. Its stated aim is to let organizations run analytics and machine learning workloads on sensitive data without exposing individual records to any single party.

The company's product line includes PrivateSQL, privacy-preserving AI tooling, and an encrypted smart contract platform. It has also been described as a privacy-first cloud computing platform built on blockchain technology, intended to support cloud-scale applications with decentralized trust and privacy protection. Reported collaborators include Google Cloud, Meta, BMW Group and Equifax.

A representative deployment is a system built with Meta for measuring fairness and possible bias in AI models. The design combines highly sensitive demographic attributes held by one party with model classification and regression outputs held by another, using multi-party computation, homomorphic encryption, differential privacy and zero-knowledge range proofs, so that user data is not revealed and measurement results are visible only to the organization that owns the models. The system is described as operating at the scale of millions of users.

Founding story

Oasis Labs was founded in 2018 by UC Berkeley professor Dawn Song, who serves as Founder and CEO, with co-founders Bobby Jaros (COO), Noah Johnson (CPO) and Raymond Cheng (CTO).

Business model

Oasis Labs supplies privacy-preserving data and AI technology to organizations that need to compute on sensitive data, delivering it through products such as PrivateSQL, privacy-preserving AI tooling and an encrypted smart contract platform, and through joint engineering engagements with large enterprises and platform partners.

Traction

Reported working relationships with Google Cloud, Meta, BMW Group and Equifax; a fairness-measurement system built with Meta operating at the scale of millions of users; and a team of about 40 employees across San Francisco and New York.