Scholarly search · 2017–2019
Delve
I built Delve as an academic search engine for dataset retrieval and document analysis, initially targeting data mining and machine learning researchers. The system helps users locate relevant datasets for specific topics, analyze how datasets have been used across papers, inspect citation relationships, and upload scientific PDFs for visual analysis of citation networks. The goal was ambitious but focused: make dataset discovery easier and expand the system to more research fields over time.
scholarly searchcitation networksdataset retrievalsucessful





Outcome
Turned dataset discovery into an interface problem: helping researchers locate benchmark data and understand how it is used.