26 April, 2024INVEMA, technological unit of AFM Cluster, is one of the participants of the ADAPT-IA project, funded by the Basque Government and SPRI through the ELKARTEK programme. The project, led by Vicomtech, focuses on the research and development of artificial intelligence applied to language technologies in Basque. The UPV/EHU, the Spanish Railway Industry Association MAFEX, AIC-Automotive Intelligence Center, and the Terminology and Lexicography Centre UZEI also participate in the project.
The main objective of the project is to enhance and naturalise the integration of Adaptive AI and Machine Learning Operations techniques in the industrial processes of different Basque sectors, exploring methodologies that optimise their maintenance, adaptation and continuous deployment of neural models in production. All of this combining both the opportunities offered by the potential of AI and the creativity and knowledge that the human mind possesses.
Three main lines of research are worked on, among which are the generation, compilation and standardisation of data by industrial sector; Adaptive AI technologies and methodologies and MLOps (Machine Learning Operations), and finally AI Modelling of Language Technologies (TL).
In addition, some of the potential applications of the project include, for example, the extraction and structuring of information (reports, calls, emails, etc.) to improve internal organisation and with the client; the generation of new multilingual content (technical manuals, documentation, resources, videos, etc.); or the improvement of the provision of assembly, inspection and maintenance services for installations.
ADAPT-IA enables a qualitative leap in the development of Adaptive AI paradigms in Language Technologies in Basque enhanced with Machine Learning Operations practices, significantly improving the current processes of data curation and structuring in Basque industrial companies.
