Skip to content

ARC

Designing for strategic exploration and responsible AI in systematic literature reviews.

ARC interface overview: search bar, search history with comparisons, search results table, and LLM-powered relevance filtering.

Systematic literature reviews are fundamental to scientific progress, but the work is spread across a fragmented tool ecosystem that imposes a high cognitive load and suppresses the iterative, exploratory side of scholarship. An exploratory design study with 20 experienced researchers surfaced three friction points: refining queries across multiple databases, the scale and pace of modern publishing, and the tension between automation and scholarly agency.

ARC is a design probe built from those findings. It brings multi-database search, a transparent record of how a query evolved, and verifiable AI-assisted screening into a single environment. In a comparative study with 8 researchers, that integration moved effort away from administrative overhead and toward strategic exploration, using external representations and visible AI reasoning to keep expert judgement in the loop.

People

Papers