Deploying AI to predict student drop-out & Used AI to teach how to deploy AI to automate processes

Actors involved

Vocational teachers, IT specialists, students, administration.

AI technologies used

  • Tool created for project purposes
  • Microsoft’s Copilot AI assistant ; ChatGPT ; tool for detecting plagiarism.

Planned activities

  • Searching for strategies to collect data from different sources for predicting drop-out risks.
  • It focuses not on theoretical teaching and learning about AI but on deploying AI in practice. One of the activities aimed to establish the sequence of actions but was less successful than expected.


  • To analyse how and if AI can predict learners’ dropout and assess if this solution suits the institution. The most significant benefit is that the warning signal comes before the staff member notices it, allowing them to make data-informed decisions on actions to avoid dropouts.
  • To suggest and apply solutions on how to predict students’ dropout rates.

Project origin

  • The head of the institution initiated the project idea several years ago based on the experiences of other institutions that were presented at international conferences. The project’s initial idea was to apply DI in some way, and then they looked for ways to make it more relevant to the issues that the institution encounters.
  • The teacher gives classes about new technologies, so allowing students to experiment with AI corresponds to technological innovations, enabling learners to practice the skills and awareness about the possibilities and risks of the AI tool Copilot.

Lesson planning

AI is a motivational tool in classes as learners are curious to explore new tools and innovations. Students want to see more possibilities of AI applications that are more than just for chatting and text generating, but even for preparing templates of instructions that might be a reasonable basis for the result.
The teacher assesses whether the deployment of AI makes it easier to find the right solutions to create a suitable sequence of action.
The teacher also teaches learners to create the correct prompts.
AI is also used for detecting plagiarism.

Technical requirements

Institution has licensed Microsoft 365; Copilot is integrated into Microsoft 365. Apart from this, no other technical requirements were needed.

External stakeholders

In the beginning, they mainly learned independently from the digital resources available, but now, many different courses are available, so they choose what they need at the time. Next to this, teachers share their experiences among themselves.

Issues encountered

  • More issues encountered from the technical perspective: AI for predicting learner attrition requires data to be collected from different sources, e.g., Moodle, e-journals, and other databases, which immediately leads to the problem of data protection and different data formats. Then, it becomes apparent that artificial intelligence is less strong on this issue than it seems.
  • There’s a lack of tools that would fit the specific needs of vocational schools. The ones that exist are expensive or have restricted access.

About Results

  • To suggest and apply solutions on how to predict students’ dropout rates.
  • The aim is straightforward: as a teacher teaches about new technologies, he wants students to practice these technologies immediately. Learners value the possibility of practicing the proper use of AI to find solutions, create instructions, and critically assess them, as well as the relevance and real-world application of this experience. The generated instructions also allow the detection of errors when instructions are generated not clearly. In this case, students must rethink how to make it better.

Links to external resources

Table of Contents

This project has been funded with support from the European Commission. This publication [communication] reflects the views only of the author, and the Commission cannot be held responsible for any use which may be made of the information contained therein.