AI for education and schools: from AI tutors to student progress analytics
Vocational, higher vocational and university institutions are handling ever more students with the same teaching staff. At the same time, the pressure to deliver personalised learning, identify drop-outs early and assess open questions more quickly is growing. Artificial intelligence gives education organisations practical tools, from AI tutoring and automatic grading to student chatbots and learning analytics, that support teachers without taking over their pedagogical role. We build AI solutions that fit your LMS, SURFconext identity and the strict frameworks of the GDPR, the AI Act and the Inspectorate of Education.
Discuss your education AI case View applicationsWhy AI in education is a logical step
An average teacher in vocational or higher vocational education assesses hundreds of open questions, essays and placement reports each year, while classes grow and the call for individual feedback grows louder. At the same time, students expect answers to their administrative questions around the clock, about enrolment, study progress, the BSA standard or placement arrangements. AI is no miracle cure, but it offers a set of tools that automate routine tasks and give teachers room for what only they can do: teaching, coaching and pedagogical judgement.
Educational institutions work with large volumes of data that often go unused: grade history in Magister, SOMtoday or Eduarte, learning activity in Itslearning, Moodle or Canvas, progress data from Numo or formative tests, and demographic data via BRON and DUO integrations. By combining these data streams in a GDPR-compliant way, models emerge that flag drop-out risk before students actually fall behind, personalise practice material to each student's level, and give teachers an evidence-based view of where remedial or catch-up lessons deliver the most value.
What sets AI in education apart from other sectors is how sensitive the target group is. A large share of students are minors, the content is publicly funded, and the assessment process directly affects qualifications with legal consequences. We build AI solutions that take this reality as their starting point: not as a compliance layer bolted on afterwards, but as an architectural principle from the very first sketch.
Core applications for vocational, higher vocational and university education
The applications vary by level of education, but the common thread is the same throughout: more tailored support for each student, faster feedback, and visible patterns where there is currently only gut feeling.
AI tutor for personalised learning
An AI tutor adapts practice material to the level, pace and error patterns of each student. Rather than a generic question sequence, it follows a path that adjusts based on previous answers. We build tutors that run inside your existing LMS via LTI 1.3 and write xAPI events back to your learning record store for analysis.
Automatic grading of open questions
Open questions, short essays and programming assignments can be pre-marked by language models against a rubric set by the teacher. The teacher sees a draft grade with justification for each assessment criterion and adjusts where needed. Official grade assignment remains with the teacher: AI provides the first reading, not the final judgement.
Plagiarism and AI-text detection
Students use ChatGPT and similar tools; that is a fact of life. Our detection pipelines combine classic plagiarism checking with indicators for AI-generated text and stylistic deviations from the same student's earlier work. Not a black-and-white verdict, but a conversation starter for the teacher.
Content generation for teachers
Practice questions, summaries, example case studies and draft exam questions can be generated by LLMs based on the syllabus and previous assessments. The teacher selects, briefly edits and validates. Again, responsibility for the content stays with the human, but the first version is ready in minutes.
Student chatbot for administrative questions
"How do I enrol for a course?", "When is the resit?", "Which documents do I need for my placement?" A chatbot trained on your study guide, examination regulations and internal FAQ answers these questions around the clock and escalates only genuinely complex cases to the student dean or student services desk.
Early warning of dropouts
By combining login frequency in Itslearning, Magister or Canvas with grade history, attendance and progress on assignments, you get an early picture of students at risk of dropping out. The study adviser receives not a black-box score but a list of the underlying signals to open the conversation.
Dutch as a second language and language learning support
For international students or NT2 (Dutch as a second language) programmes, adaptive language models offer real-time speech and text correction, pronunciation analysis and level-appropriate practice material. Useful at ROCs with language bridging programmes, international courses and civic integration.
Study progress analytics
Dashboards that show teachers and programme managers which learning objectives the group has mastered, where there is wide variation and which topics need revisiting. Aggregation at class, year and programme level, with drill-down to the individual student where the teacher deems it necessary.
SURFconext and ECK integration
Our AI components sign in via SURFconext, so single sign-on and federated identity remain intact. For digital learning materials, we connect to the Educatieve Contentketen (ECK, the Dutch Educational Content Chain) so that licences, deliveries and usage are recorded consistently.
How Appfront builds AI for educational institutions
We don't build generic education SaaS. Our approach starts with your existing infrastructure, whether that is Magister, SOMtoday, Eduarte, Itslearning, Moodle or Canvas, and connects AI components through standards such as LTI 1.3, xAPI, OneRoster and the SURFconext federation. There is no parallel system for teachers to learn separately; instead, we enhance what they already use.
The ICT coordinator at a vocational college with nine campuses has different requirements from a digitalisation manager at a university of applied sciences with its own IT department, or a university faculty already taking part in a SURF pilot. We tailor scope, governance and technology choices to your situation: which data is available via DUO and BRON, how mature your LMS data practices are, which data protection officers need to be involved, and which decisions still have to be taken by the Executive Board.
Our way of working is iterative and informed by pedagogy. We start with one concrete use case, such as a study advice chatbot or an AI maths tutor, build it first as a proof of concept, validate it with a small group of teachers and students, and only expand once the pedagogical and technical results justify it. No large-scale tender for a platform that may not fit your educational vision.
From first conversation to working educational AI
Our approach to AI projects in education follows four phases. Each phase delivers a concrete, testable result, with no year-long analysis exercise that teachers or students never notice.
1. Pedagogical and technical assessment
We map your educational vision, data landscape and legal framework: which data sits in Magister, SOMtoday, Itslearning or Canvas, which data processing agreements are in place, and what arrangements exist with the participation council and the Executive Board.
2. Proof of concept
Within a few weeks, we build a working prototype of the chosen use case on a protected dataset. This could be an AI tutor for one subject module, a chatbot covering part of the study guide, or an early-warning model for one programme.
3. Integration and pilot
The validated model is connected to the LMS, SURFconext and, where relevant, ECK and BRON via LTI 1.3 or API integration. We run a pilot with a defined group of teachers and students and measure impact and acceptance.
4. Scaling and monitoring
After a successful pilot, we scale up to a wider rollout, with ongoing monitoring of model performance, bias and data drift. We retrain when new academic years or curriculum changes call for it.
Technology we use for educational AI
The choice of technology depends on the application. For adaptive tutors, we use knowledge tracing models and transformer-based dialogue systems. For automatic grading, we combine rubric-based evaluation with large language models, and always keep the teacher in the loop. For learning analytics, we work with classical statistical models and gradient boosting, since a well-founded regression model often matches or outperforms a deep neural network in predictive value, and it is far easier to explain.
We connect AI components through open education standards rather than proprietary integrations. LTI 1.3 for embedding in the LMS, xAPI for learning activity events, OneRoster for course and user data, and SURFconext for identity and single sign-on. This keeps your institution independent: the AI component can be replaced in future without teachers or students noticing any change in the user experience.
Test your idea first: a working prototype in 1 day
With OneDayBuild, we turn your idea into something tangible in one day for €1,150, so you can see whether further development is worth the investment. Decide to go ahead with the full build? Then we credit the full cost.
Explore OneDayBuild →Education data, GDPR and the AI Act: what you need to know
Educational institutions process special category personal data, often of minors, with a direct impact on students' educational paths. Our AI architectures take the GDPR, the AI Act, the SURF Normenkader and the guidance of the Dutch Inspectorate of Education as their starting point, not as a compliance checklist added afterwards.
GDPR and minors
For students under sixteen, many processing activities require consent from a parent or guardian. We build consent flows that respect this age threshold, with clear information and the option for parents to object or request access. Data subject rights are technically implemented, not just documented on paper.
AI Act: high-risk classification
Under the EU AI Act, AI used in education and vocational training is in part classified as high-risk, particularly for admissions, assessment and proctoring. Article 5 also prohibits certain forms of emotion recognition in educational settings. We assess every use case against the AI Act framework and document risk assessment, data quality, human oversight and transparency as required.
SURF Normenkader and the Inspectorate of Education
The SURF Normenkader for information security and privacy, and the guidance of the Inspectorate of Education, are binding reference points for vocational, higher vocational and university education. Our solutions align with these frameworks: pseudonymisation choices, logging, audit trails and retention periods are made explicit and are auditable.
Education pseudonyms law and BRON data
Data exchange with DUO and BRON is subject to specific rules on education numbers and pseudonyms. AI models trained on these datasets must meet strict minimisation requirements. We separate training data from production data, use pseudonymous identifiers, and remove directly identifiable data from model inputs wherever possible.
No proctoring with emotion recognition
AI-based online exam invigilation is a sensitive category. Under the AI Act, emotion recognition of students is prohibited, and proctoring itself falls under high-risk AI. We generally advise against automated proctoring models and, where needed, look for solutions that support human oversight rather than replace it.
Hosting and data sovereignty
Education data remains within the EU, preferably on Dutch infrastructure. We work with SURF-approved cloud providers, our own on-premises deployments and privacy-friendly model hosting. No uncontrolled exchange with US-based LLM APIs for student or staff data.
Practical scenarios for vocational, higher vocational and university education
AI in education is no longer a distant prospect. These are realistic projects that connect to existing systems and frameworks, and that can be operational within one or two academic years.
Early detection of dropout at a vocational college
A vocational college with 18,000 students sees structural dropout in the first year. By combining grade history from Eduarte, login frequency in Itslearning and attendance data, a dashboard is created for study guidance counsellors. Not a score that "writes off" students, but a list of signals, for example a student who hasn't logged in for three weeks and is declining in one subject, which the counsellor can use to start a conversation.
AI maths tutor in higher vocational education
A university of applied sciences with a highly diverse maths intake is building an AI tutor that runs inside Canvas via LTI 1.3. The model adapts practice material to each student's level, provides explanatory videos for specific error patterns and reports aggregated progress to the lecturer. On the dashboard, the lecturer can see which topics the group has not yet mastered and adjust the lecture accordingly.
Student chatbot for a university faculty
A university faculty receives thousands of emails every year about OER, BSA, resits and internship procedures. A chatbot trained on the study guide, OER and internal FAQ, with SURFconext login so the bot knows which programme and cohort the student belongs to, answers standard questions 24/7 and escalates genuinely complex cases to the study adviser.
Automatic grading of programming assignments
A computer science degree programme at a university of applied sciences has hundreds of programming assignments to mark every week. An AI pipeline runs the code against test cases, assesses code quality against a fixed rubric and produces a provisional grade with justification. The lecturer reviews the output and makes adjustments, so marking time per assignment drops considerably while responsibility for the substance stays with the lecturer.
NT2 support for civic integration
A vocational college running language bridging programmes uses an adaptive speaking and writing coach for NT2 (Dutch as a second language) students. Students practise pronunciation and phrasing at their own pace, receive immediate feedback and follow a personal learning path. The lecturer uses aggregated data to focus classroom teaching on the points the whole group finds difficult.
Content generation for an exam question bank
A programme team wants a larger practice and exam question bank without individual lecturers spending hundreds of hours writing questions. An LLM generates draft questions based on the syllabus and previous exams. The subject group reviews, edits briefly and validates them, so the question bank grows faster while content quality assurance is maintained.
Why Appfront for AI in education
Education-specific architecture
We understand the difference between an SIS, an LMS and a learning record store. Between DUO integrations, BRON data exchange and SURFconext federation. Between formative and summative assessment. We translate that domain knowledge directly into AI models and integrations that fit your landscape.
GDPR and the AI Act from day one
Compliance is not an afterthought but a design principle. Risk assessment, data minimisation, human checkpoints and explainability are built into the architecture, not added as documentation afterwards. Your Faculty Board and Executive Board receive technical answers to their questions, not only legal ones.
Open standards and no lock-in
LTI 1.3, xAPI, OneRoster and SURFconext: we integrate through the standards your institution already invests in. That means AI components are replaceable and your data landscape isn't locked into proprietary formats from a single vendor.
Frequently asked questions about AI in education
Interested in using AI in your education organisation?
Discuss your case with us. Together with your ICT coordinator or digitalisation manager, we will map where AI delivers the most value, both pedagogically and technically, and which compliance requirements are decisive. No obligation.
Schedule a conversation