What is BlindAI?
BlindAI is a fast, easy-to-use, and confidential inference server, allowing you to deploy your model on sensitive data. Thanks to the end-to-end protection guarantees, data owners can send private data to be analyzed by AI models, without fearing exposing their data to anyone else.
We reconcile AI and privacy by leveraging Confidential Computing for secure inference. You can learn more about this technology here.
We currently only support Intel SGX, but we plan to cover AMD SEV and Nitro Enclave in the future. More information about our roadmap can be found here.
Our solution comes in two parts:
- A secure inference solution to serve AI models with privacy guarantees.
- A client SDK to securely consume the remote AI models.
- BlindAI server: 0.4.0
- BlindAI client: 0.4.1
- Simple and fast API to use the service
- Model and data protected by hardware security
- Support of Remote Attestation with TLS (DCAP library)
- Easy to install, deploy, and maintain
What you can do with BlindAI
- Easily deploy state-of-the-art models with confidentiality. Run models from BERT for text to ResNets for images, through WaveNet for audio.
- Provide guarantees to third parties, for instance, clients or regulators, that you are indeed providing data protection, through code attestation.
- Explore different scenarios from confidential Speech-to-text, to biometric identification, through secure document analysis with our pool of examples.
What you cannot do with BlindAI
- Our solution aims to be modular but we have yet to incorporate tools for generic pre/post processing. Specific pipelines can be covered but will require additional handwork for now.
- We do not cover training and federated learning yet, but if this feature interests you do not hesitate to show your interest through the roadmap or Discord channel.
- The examples we provide are simple and do not take into account complex mechanisms such as secure storage of confidential data with sealing keys, an advanced scheduler for inference requests, or complex key management scenarios. If your use case involves more than what we show, do not hesitate to contact us for more information.
Who made BlindAI?
BlindAI was developed by Mithril Security. Mithril Security is a startup focused on confidential machine learning based on Intel SGX technology. We provide an open-source AI inference solution, allowing easy and fast deployment of neural networks, with strong security properties provided by confidential computing by performing the computation in a hardware-based Trusted Execution Environment (TEE) or simply enclaves.