Providing models, code and data in order to make Riksarkivet's archival holdings more useful and used.
The Swedish National Archives (Riksarkivet) primarily manage government records for future generations. Access to cultural heritage and authentic information that can be used for research, judicial matters, and public transparency is fundamental to our democratic society.
Riksarkivet has identified a significant opportunity to establish an AI lab, known as AIRA (AI at Riksarkivet), to focus on research and development as well as product development within Riksarkivet. As a lab within a supervisory authority, we offer services, datasets, and models related to our public mission.
We work primarily with Handwritten Text Recognition, semantic search, and to some extent, Large Language Models (LLMs). We share our training data and models on platforms like Huggingface 🤗 and elsewhere. If you are interested in improving or utilizing our models, we would be delighted to collaborate!
Find us here 👉
