Projektstatus: Idea Release In use

QLever Federation

QLever Federation reduces redundant data storage and energy use in research by enabling direct, remote querying of massive datasets—paving the way for more sustainable data infrastructure.

#Round 4

QLever is an open-source knowledge graph engine designed for speed and scale, capable of handling billions of facts with full-text and spatial search. It is already used by projects like UniProt and OpenStreetMap to make complex data accessible and explorable.

With QLever Federation, we are tackling the next big challenge: enabling researchers and institutions to query massive datasets directly, without copying them locally. This avoids redundant data storage, reduces network traffic, and significantly cuts energy use. If adopted widely, this approach could save up to 7.2 GWh per year, enough to power 2,400 Swiss households.

How we will achieve real-world impact

By reducing the need to duplicate and host large datasets in multiple places, QLever Federation addresses a hidden inefficiency in research infrastructure: the digital equivalent of unnecessary commuting. During the prototyping phase, we will measure energy savings and explore generalizable patterns to guide wider adoption of sustainable, federated knowledge graphs.

Team

Adrian Gschwend
Johannes Kalmbach
Hannah Bast
Ludovic Muller

Contact and Website

PROJEKTÜBERSICHT