mithril-security/blindai
Confidential AI deployment with secure enclaves :lock:
Leverages Intel SGX trusted execution environments paired with the Tract inference engine to run ONNX models within hardware-isolated enclaves, preventing server operators from accessing user data or model internals. Supports two deployment modes: querying pre-hosted models via a Python client API, or self-hosting a BlindAI Core server instance to deploy proprietary models with confidentiality guarantees built into the infrastructure layer.
512 stars. No commits in the last 6 months.
Stars
512
Forks
35
Language
Rust
License
Apache-2.0
Category
Last pushed
Mar 19, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/mithril-security/blindai"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
ICME-Lab/jolt-atlas
Fast zkVM born at a16z Crypto substantially adapted by ICME Labs (NovaNet) for verifiable...
socathie/circomlib-ml
Circom Circuits Library for Machine Learning
gizatechxyz/LuminAIR
A zkML framework for ensuring the integrity of computational graphs using Circle STARK proofs
mithril-security/bastionlab
A simple framework for privacy-friendly data science collaboration
ldsec/drynx
Decentralized, Secure, Verifiable System for Statistical Queries and Machine Learning on...