A new system of instructional design
Our R&D team has been working for a long time to create a Generative AI-based instructional design system.
The project concerns the implementation of a software that uses Large Language Models (LLM), with the dual objective of:
- automate course writing,
- optimize the training process.
The idea is to significantly reduce the production time of educational content, while maintaining a high quality and customization of training documentation.
What it means to introduce artificial intelligence into the training process
In practice, the effort is aimed at making the content analysis activity and the preparation of the basic project documents (macro-planning of the teaching objectives) more effective, to be then finalised in order to make them effective for the transformation into products of digital learning.
The challenge in introducing artificial intelligence is to insert a further tool to support this design phase, which not only allows for better management of the time used, but also the maintenance of high quality design standards and an exact representation of the contents.
All this with an excellent level of Supervision of the result product.
A project currently already in the testing phase
The project is being implemented with the collaboration of GRID+, startup of theSapienza University of Rome.
This involves the development of a tailor-made solution based on Generative AI, which he uses LLM with Retrieval-Augmented Generation (RAG) methodology to generate complex and personalized texts for each course.
The system is hosted entirely on Piazza Copernico's proprietary servers, which do not depend on external APIs, thus guaranteeing efficiency and ensuring maximum confidentiality in the management and transmission of information.
The solution is currently being tested, and the impacts on the work process are being experimentally measured internally.
Some significant benefits have already been perceived:
- the possibility of designing through data interrogation, which allows access to information sources to verify the effectiveness of human-machine interaction;
- the advantage of allowing the instructional designer to focus more on teaching strategies, and less on the generation of materials which is largely automated.