We're a CS theory (security/AI) research team aimed at designing and engineering new verifiable computation protocols for use on securing the training and inference of machine learning models in the cloud.
We engineer our verifiable computation protocols from the ground up to optimize them for securing cloud-based machine learning.
We've published two papers chalk-full with theory, proofs, and empirical results, with more on the way.
Scale To Production
Our ultimate goal is to scale these protocols from theory to production-level environments.
- 5,000+ Lines of Code
- 2 Verifiable Computation Protocols
- 17,000+ Paper Word Count