โ† All projects
Systems ยท DevOps

Reproducibility for Research Software

Making published research runnable again

Research code is notoriously hard to re-run: missing dependencies, drifting environments, undocumented setup. This project used containerization to replicate research artifacts and standardize the environments they need.

Type
Systems / tooling
Core tool
Docker
Platform
Linux
Goal
Reproducibility
Reproducibility for research software

01 Overview

The aim was to take research software and make it reliably reproducible by packaging it and its dependencies into Docker containers, so anyone could replicate the artifacts and results in a standardized, self-contained environment instead of fighting setup.

02 What I worked on

  • Containerized research artifacts and pinned their dependencies for repeatable runs.
  • Standardized environments so results replicate across machines.
  • Reduced the setup friction that usually blocks reuse of research code.

More detail available on request.