OpenAIRE Evidence Ledger is a small, dependency-free Python tool that turns an OpenAIRE Graph search into a human-reviewed evidence pack. It is for researchers, public-sector innovation teams and open-science practitioners who need a shortlist they can inspect rather than a fluent answer whose factual boundary is unclear.

Code

Video

Raw scholarly metadata is noisy and unclassified. Training AI tools that distinguish a methodology paper from a dataset paper from a review requires labelled data, and hand-labelling thousands of papers is slow, expensive and hard to scale across disciplines. We built a complete feedback loop that turns the OpenAIRE Graph, a vast but unlabelled index of research outputs, into validated training data for AI classifiers, at scale and with minimal human effort.

Code

ScholarMind is an AI-based research and knowledge management system designed to help researchers organize, connect, search, and reuse scientific knowledge. It provides a structured environment for managing research information and developing research memory rather than treating scientific papers as isolated documents.

Video:

Code

Archive

What research connections—and gaps—link mathematical and model-based reasoning with AI literacy, particularly in relation to interpretation, uncertainty, evaluation, and judgment? The question matters because AI literacy frameworks increasingly require people to “critically evaluate” AI, yet this phrase does not by itself specify the cognitive operations involved when an AI output is uncertain, probabilistic, or potentially wrong. The exploration therefore examined whether AI literacy research already connects to the traditions that study reasoning under uncertainty and to empirical research on when humans appropriately rely on, override, or reject AI advice.

Code

Archive

Replication Radar answers that on the OpenAIRE Graph. Search a field and, for each high-impact claim, it shows whether anyone has independently proven it: a signed Science Live replication verdict (validated / contested / refuted), read live from the nanopublication network, author-agnostic. Where the Graph knows only cited / not-cited, the Radar adds checked / not-checked.

Video

Code

Archive

Worked example

OpenPhys V2 is an agent-driven, open-source computational physics engine that solves Partial Differential Equations (PDEs) using Physics-Informed Neural Networks (PINNs). Traditional numerical PDE solvers (FEM/FDM) require deep domain expertise, manual mesh generation, and rigid code bases. Conversely, standard machine learning approaches lack physical grounding and academic context. 

Code

Data Footprints is a reusable analytical workflow and interactive application that uses the OpenAIRE Graph to make dataset citation pathways visible and interpretable for responsible research assessment. Starting from a researcher’s ORCID, the application discovers their datasets through the OpenAIRE MCP connector powered by Alien Intelligence, validates authorship, reconciles multiple records that may refer to the same conceptual dataset, and traces publications that cite those datasets through OpenAIRE Graph API V3. It then compares the dataset creators with the authors of each citing publication.

Code

Story

Funders can count grants and publications. Can they also trace the public evidence to datasets and software, identify which persistent links are missing, and measure whether those links improve over time? Open Quantum Evidence Atlas is a working, no-login Evidence Chain Auditor for funders, programme managers and Open Science teams. It answers one operational question: after research is funded and published, can the public evidence be followed to datasets and software—and which persistent links must be fixed when it cannot?

The full meaning of D-RectifyL is Deterministic Rectification Layer; an interactive, open-source tool that folds any protein or DNA sequence in microseconds using a physics framework derived entirely from open-access Zenodo preprints indexed in the OpenAIRE Graph and the EOSC community. Its a single-cell Google Colab application with a widget interface. The user types any UniProt accession, gene name, or raw sequence. The tool fetches live biological data from UniProt, queries the OpenAIRE Graph for the underlying physics corpus, and returns: the dominant phosphorylation site (the e₄ snap node), fold class, PTM site ranking, OZIP 8-component state, speed benchmarks, and a full biological identity card — all in under 30 milliseconds.

Notebook

Video

Code

Archive

The question was motivated by a simple problem: “the future of work” is not a single research field. It cuts across economics, sociology, management, technology, labour studies, organisational research, public policy and, increasingly, artificial intelligence. A conventional literature search can retrieve relevant publications, but it is much harder to understand how these publications are connected, which themes are emerging, which researchers and organisations shape the field, and how research priorities change over time. We therefore asked whether the OpenAIRE Graph, accessed through the Alien Intelligence MCP connector, could be used not merely to find papers, but to explore the structure and evolution of a research landscape and turn the resulting evidence into an intelligible research narrative.

OpenGap is an evidence-first web application that helps researchers move from a broad research idea to an inspectable potential-gap hypothesis. Instead of returning another long list of papers, it compares several kinds of research activity available through the OpenAIRE Graph — publications, funded projects, software and datasets — and surfaces simple structural imbalances that may deserve deeper investigation.

Code

Empirical finance contains thousands of published claims about how prices move. A person testing them on a private dataset normally chooses which claims to test, often with some intuition about what the data may say. The result can look like an independent test while behaving like a search for a convenient result. We asked whether a public research graph could make that choice explicit and auditable.

Code

Story