Research datasets and code repositories your AI assistant can search
Finding a reusable dataset or the code behind a paper is often harder than finding the paper. Kahubi connects to Zenodo and DataCite, which together cover tens of millions of DOIs for datasets, software and other research outputs, to Hugging Face for machine-learning datasets and models, and to Stack Exchange for methods and coding questions.
The assistant returns DOIs, licences, versions and file lists, which is exactly what a data availability statement or a methods section needs. Community answers from Stack Exchange are marked as advice rather than peer-reviewed sources.
4 sources you can connect
- DataCiteDOIs for research datasets, software, samples and other non-article outputs.Hosted by Kahubi
- ZenodoCERN’s open repository for datasets, software, papers, posters and supplementary material.Hosted by Kahubi
- Hugging Face HubSearch models, datasets, Spaces and papers on the Hugging Face Hub.Official server
- Stack ExchangeQ&A across Stack Overflow, Cross Validated, MathOverflow, Academia and 170+ sites.Community server
Questions researchers ask these sources
“Find published datasets on Baltic Sea eutrophication with a DOI.”
DataCite records with creators, repository, year and DOI, ready to cite in a data availability statement.
DataCite“Is there a public dataset of Swedish parliamentary speeches on Zenodo?”
Records with title, creators, date, DOI, licence and the file list with sizes.
Zenodo“Find open datasets for Swedish speech recognition and summarise their licences.”
Dataset repositories with their cards, sizes and licence fields.
Hugging Face Hub“How do I handle repeated measures with missing data in a mixed model? What does Cross Validated recommend?”
Relevant threads with the accepted and top-voted answers, marked as community advice rather than peer-reviewed sources.
Stack Exchange
Frequently asked questions
How do I cite a dataset?
Can it check a dataset’s licence?
Can it analyse a dataset it finds?
Browse other sources
Search datasets & code sources in one research chat
Free to start. Connect a source in one click and get answers that cite the records they came from.