Mark Zakhvatkin and Quintus Stierstorfer are pioneering figures in educational AI at IU International University of Applied Sciences. Mark, as Director of AI & Data, brings his unique blend of subnuclear physics background and McKinsey consulting experience to spearhead AI innovations in higher education, while Quintus leverages his mathematics expertise to lead the "Synthetic Teaching" department, focusing on GenAI applications. Together, they are driving the development of Syntea, an AI-powered learning assistant that has already demonstrated significant impact on student performance, including a 27% decrease in course completion time and handling over 33,000 course-related questions monthly.

In November 2024, LIBF and IU Group’s learning initiative won the Innovation in Higher Education & Learning award at the MERIT Summit & Awards in Dubai.

In what ways does Syntea adapt its learning approach for students with diverse learning styles, backgrounds, and academic preparedness levels, ensuring an inclusive support experience?

Syntea personalises its learning approach by adapting to each student's individual pace, knowledge level, and specific needs. Through continuous tracking of progress, it identifies strengths, weaknesses, and knowledge gaps, allowing students to focus on the areas that require the most attention.

The platform provides instant, targeted feedback, helping students prepare effectively for exams while promoting critical thinking and deeper understanding of the material.

Additionally, Syntea integrates seamlessly into IU Group’s learning ecosystem, such as the myCampus app, making it accessible at any time and place. This flexibility supports a continuous learning process without the need for extensive materials or complex platforms. By reducing bias in the feedback process, Syntea further enhances usability for people from diverse backgrounds, ensuring its inclusivity. By tailoring learning paths, enabling selective lesson focus, and aligning content with students’ academic preparedness, Syntea ensures an inclusive and effective educational experience for a diverse student body.

How did you address potential concerns about data privacy and ethical AI use while developing a personalised AI study assistant?

While developing Syntea, IU Group prioritised data privacy and ethical AI use by adhering to strict compliance with GDPR regulations and implementing robust data protection measures. Personal data is handled with transparency, ensuring that students are informed about data usage, for example, there is a disclaimer on the start-up of Syntea. Ethical considerations were integral to the development process, with Syntea designed to operate within clear boundaries. So-called guiding rails are implemented to ensure that conversations with Syntea remain focused and do not veer off-topic. This means users are guided away from non-study-related discussions and kept on track within their intended learning path. This is essential for maintaining focus, preventing misunderstandings, and ensuring a positive user study experience.

Can you elaborate on the specific machine learning and natural language processing techniques used to create Syntea’s personalised tutoring?

Syntea's conversational capabilities are powered by cutting-edge generative language models, leveraging state-of-the-art Large Language Models (LLMs) provided by platforms like OpenAI/Azure and Anthropic's Claude, as well as a few LLMs that were fine-tuned with a proprietary IU Group data. These models enable Syntea to understand natural language inputs and engage in meaningful, context-aware conversations with users.

We are also actively experimenting with the AI agentic architectures and are very excited to hear students’ feedback on the prototypes we’ve prepared for them built in the new architectural paradigm.

The underlying technology and logic driving Syntea, along with its front-end services, have been thoroughly developed in-house by IU Group to meet the specific needs of its students. Syntea features continuous learning capabilities, enhancing its responsiveness and relevance. For instance, it retains the context of previous interactions, allowing it to reference earlier questions and answers, ensuring a more personalised and seamless learning experience for each student.

What measurable impacts have you observed in student performance, engagement, and learning outcomes since implementing Syntea?

Since implementing Syntea, we have observed significant improvements in student performance, engagement, and learning outcomes. A first scientific analysis conducted by IU Group revealed that the average course completion time for online students has decreased by 27%, highlighting Syntea’s ability to enhance learning efficiency. Additionally, 60% of students who assessed their knowledge before starting a course reported that they could skip certain units, allowing them to focus on areas where they truly needed to improve.

Engagement levels have also been remarkable with more than 800 students actively engaging with it daily. To date, Syntea has exchanged over 8 million messages with students, and it handles 33,000 course-related questions every month.

This high level of interaction demonstrates how Syntea not only supports learning but also fosters a more dynamic and responsive educational environment.

How do you envision the future evolution of AI study assistants, and what potential challenges or opportunities do you anticipate in scaling such learning technologies?

The future evolution of AI study assistants like Syntea lies in their ability to become even more personalised, context-aware, and seamlessly integrated into diverse learning environments.

By making Syntea seamlessly available in IU’s online learning platform myCampus, we made it universally accessible, for example.

Furthermore, remarkable advances in AI, that can be observed particularly in generative models, will enable these tools to provide highly customised learning experiences tailored to individual needs, learning styles, and career goals. With natural conversational interfaces, Syntea can deliver more intuitive, human-like interactions, helping learners engage more deeply and fostering better understanding. Additionally, Syntea can shift from summative assessments, which focus on end results, to formative assessments that emphasise continuous learning and improvement, supporting students as they progress their learning goals.

We envision AI study assistants not only as tutors but also as lifelong learning companions, offering real-time feedback.

Do you have an initiative in organisational, leadership, and people development worth sharing? The next edition of the MERIT Summit & Awards is coming up soon on 5-6 June 2025 in Lyon. Submit your application until 7 March – shortlisted initiatives will be invited to present at the upcoming summit in Lyon.