The human body is complex. AI and data may be used to solve these complexities and improve health outcomes.  To drive more informed data-backed decisions, DeepHive introduces, DeepHuman, a proprietary AI application for our protocol.


DeepHive aims to develop an unprecedented, end-to-end, open-source AI application layer for our protocol that combines the best qualities of humans and machines. Our innovative and intuitive AI application layer, a clinical decision support (CDS) module that plans to use differentially private federated learning, aims to understand relations between various biomedical, nutrition and wellness concepts.  Providing practitioners with AI-driven diagnoses, treatments, formulations and other results in context across multiple big data sources enables healthcare practitioners to gain insights and increase discovery, potentially making more informed and data-backed clinical and wellness decisions.

Federated Learning


DeepHive plans to utilize federated learning in its proprietary AI application layer, DeepHuman. Most machine learning models only use centralized data located on one main device. Federated learning allows multiple devices to collaboratively and collectively learn under a decentralized intelligence framework while keeping all of the training data on each of the multiple devices. Federated learning can exponentially accelerate machine learning models by using the data of multiple parties.

Differential Privacy


To ensure HIPAA compliance, DeepHive plans to apply differential privacy to its proprietary AI application layer, DeepHuman.  Applying differential privacy to federated learning may allow our protocol to train machine learning models using another party’s data without the other party explicitly sharing data or knowing the other party’s data. Differential privacy is a privacy-centric model that is designed to be compliant with various healthcare and privacy regulations such as HIPAA and GDPR.