How is BlindAI structured?
BlindAI’s goal is to allow users to consume AI models without showing their data to the AI provider, thanks to Confidential Computing.
The project has been historically intended for AI engineers but has been modified to onboard developers easily with ready-to-use AI APIs.
BlindAI API is a simple client-side SDK to query popular AI models that we host inside a Trusted Execution Environment to help developers get started quickly with AI whilst keeping their data private. BlindAI API uses BlindAI Core behind the scenes.
BlindAI Core contains the server and client implementations to upload a model inside a Trusted Execution Environment and query it securely.
BlindAI API includes an open-source Python library, allowing developers to query popular AI models without needing to trust us with access to your data, thanks to Confidential Computing.
BlindAI Core facilitates privacy-friendly AI model deployment by letting AI engineers upload and delete models to their secure BlindAI server instance. Users can then connect to the server, upload their data and run models on it without giving any party access to their data.
BlindAI vs BlindAI Core: Key differences
Here is a summary of the key differences between BlindAI API and BlindAI Core:
With both BlindAI and BlindAI Core, data confidentiality is assured by hardware-enforced Trusted Execution Environments. We explain how they keep data and models safe in detail here.