What we can do for you.




Artificial Intelligence
Features that were previously impossible become possible


Intelligent Tutoring Systems
Intelligent Tutoring Systems (ITS) are computer-based systems that use AI to provide personalized feedback to students. These systems attempt to simulate the experience of working with a human tutor as close as possible.
Why?
Inside a traditional classroom, teachers may not have the time or resources to provide personalised instructions to each student. This means that students can not realise their full potential.
How?
ITS provide personalised guidance tailored to the student's individual learning styles. Instead of treating all students as interchangeable, these systems empower teachers to address each student individually which results in a learning experience which is both more pleasant as well as more efficient.
What?
We apply both classical machine learning (like decision trees or hidden Markov models) as well as deep learning algorithms to achieve the individuality needed for ITS. Thanks to our excellent background in mathematical modelling, we also have the capability to engineer custom probabilistic models tailored to your specific needs.


Natural Language Processing
Natural Language Processing (NLP) is a branch of AI that deals with techniques and algorithms to analyse, understand, and generate texts written in natural languages.
Why?
When working on certain tasks (like math exercises) a student can simply check his solution against a sample solution. However for domains like language learning this is not possible - the student has to wait until a teacher gives him feedback.
How?
We construct individualised validation components that provide personalised feedback for complex text exercises. Through this approach, students are able to make immediate improvements, leading to an accelerated rate of learning.
What?
We know how to approach common NLP tasks like text classification, morphological analysis, text parsing, text similarity evaluation. Thanks to our experts with backgrounds in both classical linguistics as well as deep learning we can engineer custom NLP components that will satisfy your unique needs.


Multimodal interaction
Multimodal interaction in education refers to the use of multiple forms of media, such as text, audio, video, and images, to enhance the learning process of students.
Why?
When acquiring new vocabulary, it has been observed that students tend to learn most effectively when they are able to articulate the new terms. However, this proficiency can only be verified by a teacher, and the student is unable to independently practice this skill.
How?
Our company develops systems that are capable of validating multimodal inputs provided by students, whether it be through oral communication via microphone or through the creation of visual images.
What?
Drawing on experience from areas like computer vision and automatic speech recognition we build systems which can work with student inputs beyond raw text or numbers. We apply state of the art convolutional and recurrent neural networks as well as transformer - based architectures that improve with each user interaction.