SciDraft: Collaborative Notebook for Lab Teams Built on Git

Scientists co-authoring papers still fight two chronic problems: version chaos across Word docs, Overleaf projects, and shared drives, and the disconnect between the manuscript and the underlying data, code, and citations. Tracking who changed what, reconciling reviewer comments, and keeping figures in sync with the latest analysis run consumes real time and introduces errors. SciDraft is a version-controlled collaborative writing environment for lab teams built on Git. It embeds live data, citations, and peer comments directly in the writing surface, so a figure can update when its data source changes, references stay in sync with a shared library, and every edit carries a Git commit history. The solution combines the rigor Git already provides for code with the ergonomics scientists expect from a document editor, hiding raw Git plumbing behind a comment-and-review interface that PhD students and PIs can use without command-line skills. An MVP would be a browser-based editor where a lab creates a project (a Git repository under the hood), writes in a structured document, links data blocks and citations, and leaves inline peer comments that map to commits and branches. Revenue is a per-seat or per-lab SaaS subscription, natural for research groups that already budget for tools like reference managers and Overleaf. Go-to-market starts inside the communities already asking these exact questions: Stack Exchange threads on using Git to collaborate with non-version-control users, keeping Zotero collections in sync via Git, and how DVCS can help scientific programming. Growth loops come from co-authorship itself: every paper drafted in SciDraft invites collaborators from other labs, and published papers can link back to their SciDraft project as a reproducibility artifact, seeding awareness across the field.