Alexander J. Schiller, PhD

The Center for Open Science is a non-profit startup with the goal of increasing openness, integrity, and reproducibility of research. I started at the Center as an intern in 2013 while working on my PhD. I signed on as a developer in 2016 and am now a tech lead in charge of three services: Preprints, Registries, and Peer Review.

The Center for Open Science current hosts 22 preprint providers. Providers are primarily topic (e.g., Psychology, PsyArXiv, 2277 preprints) or region (e.g., Indonesia, INA-Rxiv, 3214 preprints) focused. My current work focuses on building out features to improve impact, sharing, and useability while also doing application maintenance. One recent major maintenance task involved upgrading to Ember 2.18 (from 2.11) and adding stricter linting. These upgrades touched 1/4 of the preprint repo's code and contained over 6000 lines of modified code.

One fun project I enjoyed was writing a custom sitemap generator for the Open Science Framework. We required some custom site routing that made out of the box solutions less than optimal. In production, the script runs nightly on a celery beat job and ships the data to an Amazon S3 bucket (using boto3). The sitemap was needed to allow Google Scholar to index our preprints. One interesting side effect was that the sitemap script also served as a coal mine canary and alerted us on more than one occasion to issues in the production database.

My first project at the Center for Open Science was the Modular File Renderer. Scientists upload various file types to their projects. Some file types are trivial to render to HTML, but many require integrating with libraries or writing custom rendering scripts. The goal of the Modular File Renderer was to write as many renderers as possible (I wrote the first 20). The structure of the code has changed in the last few years, but I can see still bits and pieces of the original code I wrote.