DataPledge

DataPledge is a web app (with a lightweight desktop uploader) that helps researchers publish “good enough” open datasets fast, without pretending to replace full data management platforms. It guides users through a minimal, standardized checklist (license, consent/PII risk, variable dictionary, README, provenance) and auto-generates a clean dataset package ready for Zenodo/OSF/Figshare upload. An AI assistant suggests metadata, detects likely sensitive columns, and drafts documentation from notebooks and column names—but it never claims compliance or guarantees anonymization. The product is intentionally narrow: reduce the friction between ‘I should share’ and ‘it’s actually posted’ with clear, auditable steps and export formats. Teams can reuse templates per lab, track what’s been shared, and produce a simple “open data statement” for manuscripts. If users want enterprise governance, they should use heavier tools; this is the pragmatic middle ground.