Proceedings Manager

In 2021 the landscape had four problems. Expensive misfit solutions: journal platforms forced onto conference workflows, with steep learning curves and price tags, or custom systems that did not scale. Fragmented tools: three or four disconnected systems per organiser, for submissions, reviews, author communication and final preparation. Manual processing: papers by email attachment, versions in spreadsheets, review progress in email chains. Quality-control gaps: journals had verification processes, proceedings often did not, so publishers had little visibility into what they were agreeing to publish.

Proceedings Manager sat in the gap between file management and heavyweight publishing systems. I joined after its first release as Senior Product Designer and UX Researcher on a team of three backend and three frontend engineers, a QA engineer, a data engineer, a product manager, an engineering manager and a shared user researcher.

Rather than rushing to add features, we mapped the whole lifecycle on Miro boards, from paper submission to final publication. Customer success brought what users struggled with daily; sales brought what prospects asked for and what blocked deals. Five workflows needed work: submission and metadata extraction, quality checks, peer review, team collaboration and oversight, and the publisher hand-off.

August to October 2021. The platform already accepted supplementary files, but the real needs were more specific. A standalone Peer Review Statement section reflects its importance in validating proceedings. Attach Additional Files splits into files that go into the published volume (prefaces, editorial notes, volume-wide content) and copyright and legal forms (publisher consent, open-access agreements, compliance documents), with auto-matching that connects each form to its submission so nothing is missed at hand-off.

November 2021 to January 2022. Better metadata extraction was groundwork rather than a feature: it enabled validation, search and insight later. The first checks were deliberately lean and visible only to publishers and organisers. Manual-upload detection flags papers that bypassed the submission flow. Author pattern analysis flags authors with several papers in one volume, which can mean preferential selection. Citation pattern detection looks for authors who dominate a paper's references and for organisers or editors being cited, both signals of possible bias.

February to August 2022. Societies that publish several conferences a year manage reviews for hundreds of thousands of papers; matching expertise, balancing workloads, tracking rounds and keeping a record all grow with volume. Automatic assignment suggested reviewers from domain knowledge, current load and conflicts, and organisers could adjust the suggestions.

Validation is rarely one pass. A paper with three reviewers, two of whom requested changes, can go to a second round with just those two. Each round preserves previous reviews, feedback and decisions, so the history shows how a paper evolved and who took part, which proved invaluable for final acceptance decisions.

August to December 2022. The Editor-in-Chief oversees the whole proceedings but relies on domain experts for specific tracks. Section editors see only the papers in their topics, manage reviewer assignment and communication there, and make recommendations about publication. The role framework was built to extend, and reviewer pools mapped to topics followed the same structure.