The End-to-End Platform for Modular Scientific Collaboration

At a glance

Scientific research is increasingly complex and interdisciplinary. Yet the way researchers collaborate hasn’t evolved much: a paper is still treated as a single, monolithic project. This creates inefficiencies, slows down progress, and obscures the value of individual contributions.
Relink AI redefines collaboration by decomposing papers into a Directed Acyclic Graph (DAG) of modular, repo-like nodes. Each node represents a specific research task, allowing researchers to contribute at the right level of granularity—just like contributing to open-source code.

Company

Relink, Inc

Timeline

5 months

Team

4 eng, 1 PM, 2 designers

My role

Product design lead

Problem

Coarse, Opaque, and Slow Collaboration Holds Science Back

Collaboration today is too coarse-grained (whole papers), too opaque (unclear authorship), and too slow (recognition only at submission), discouraging effective participation.

HMW

How might we make research collaboration modular, transparent, and immediate?

Solution

A Clearer, Smarter Way to Collaborate on Research Papers

Through an intuitive task-graph design, Relink AI breaks down papers into modular nodes—making complex collaboration easier to navigate, contributions simpler to track, and credit fairer to assign.