ReLEAF is under active development. This roadmap summarises planned and under-consideration features of the system as well as future work of the project. Items may change based on pilot feedback, technical constraints, and governance requirements.
Platform for two-stage data sharing
AI assistant for request creation
JupyterHub instead of downloading synthetic data (to prevent misuse of synthetic data)
Outputs Registry: share publications, request code, outputs for open science
Automated synthetic data generation with differential privacy
ReLEAF Trust: transparency for stakeholders
Privacy budget management
Continual data release
Federated learning/analyitcs across multiple datasites
Feedback to schools
Semi-automate review
Standardise metadata
Share non-LEAF data (e.g. collaborate with an edtech company)
Establish governance framework (e.g. informed consents, data removal request)