How we increased the completion rate to 85% at LearnSpace

We built adaptive courses in LearnSpace that adjust on the fly to the learner’s level. The foundation is Vygotsky’s idea of the zone of proximal development: we don’t give tasks that are too easy, but we also don’t throw learners into hard mode where they want to close the tab and go watch memes.

Here’s how it works: neural network-based progress analysis spots weak points, estimates the current level, and adjusts the learning path in real time. If a topic is slipping, we add more practice and support. If a learner is moving ahead quickly, we don’t slow them down — we increase the difficulty of the scenario.

With all the hype around AI in EdTech, it was important for us not just to “add a neural net,” but to make sure learning actually gets people to the finish line. And yes, 85% completion rate is exactly the result of matching difficulty accurately.