Leipzig research centre specializing in semantic technologies, knowledge graphs, and big data integration across science, agriculture, and health domains.
INFAI is a Leipzig-based applied informatics research centre specializing in semantic technologies, linked data, and knowledge graph systems. They build tools and platforms that make large, heterogeneous datasets — from scientific publications to geospatial and agricultural data — machine-readable, interoperable, and queryable. Their practical strength lies in data integration and enrichment: taking messy, multilingual, distributed data and turning it into structured, FAIR-compliant resources usable by researchers, businesses, and public institutions. More recently, they have expanded into human-computer interaction and immersive technologies (VR/AR) applied to domains like healthy ageing and Earth observation.
Core contributor across HOBBIT (benchmarking linked data), ScienceGraph (scholarly knowledge graphs), FREME (semantic enrichment), SlideWiki (linked data for education), SLIPO (POI data integration), and SSHOC (FAIR data).
Central role in BigDataEurope (big data platform for societal challenges), QROWD (human-in-the-loop big data integration), DataBio (bioeconomy data pipelines), and HOBBIT (benchmarking data integration tools).
Contributed to SSHOC (European Open Science Cloud for social sciences), ScienceGraph (scholarly communication graphs), and RESILIENCE (research infrastructure for religious studies).
FREME developed open e-services for multilingual semantic enrichment; SlideWiki tackled multilingual collaborative courseware authoring.
e-VITA applied HCI and personalized virtual coaching for healthy ageing; CALLISTO used VR/AR and visual analytics for Earth observation data.
DataBio (agriculture, fishery, forestry data), STARGATE (adaptive microclimate management for farming), and CALLISTO (Copernicus AI services for environmental monitoring).
In their early H2020 phase (2015–2018), INFAI focused squarely on big data infrastructure and semantic content enrichment — building platforms for data integration, linked data benchmarking, and multilingual digital content services aimed at commercial validation and SME growth. From 2019 onward, their focus shifted decisively toward open science ecosystems (EOSC, FAIR data, scholarly communication) and applied AI — including knowledge graphs for research, immersive interfaces for ageing populations, and deep learning for Earth observation. This evolution shows a move from foundational data tooling toward domain-specific, user-facing applications of their core semantic and data integration expertise.
INFAI is moving from backend data plumbing toward intelligent, domain-specific applications — expect them to pursue AI-powered knowledge systems, digital twins, and immersive analytics in future calls.
INFAI operates almost exclusively as a consortium partner (12 of 13 projects), coordinating only one project (HOBBIT), which suggests they position themselves as a reliable technical contributor rather than a project leader. With 203 unique partners across 32 countries, they maintain an exceptionally broad network — they are a hub connector, not a closed-circle operator. This makes them easy to integrate into new consortia: they bring deep technical capability without competing for the coordination role.
INFAI has collaborated with 203 distinct partners across 32 countries, giving them one of the broader networks for an institute of their size. Their reach spans all of Europe and extends to Japan (e-VITA), indicating a genuinely pan-European and increasingly global collaborative footprint.
INFAI sits at a rare intersection: they combine deep expertise in semantic web and knowledge graph technologies with the ability to apply these across wildly different domains — from agriculture to cultural heritage to space. Unlike larger German research organisations (Fraunhofer, Max Planck), INFAI is lean and specialised, making them a flexible partner who can slot into consortia as the data integration and semantic technology backbone. Their track record of making heterogeneous data FAIR-compliant and machine-actionable is directly relevant to almost any data-intensive EU project.
