January, 2022
AI Assist
Smarter, stronger and simpler revision with AI
By Nicholas du Preez
The education system heavily relies on memorising information, from low-level facts in early education to complex topics and subjects in higher education. With such pressure on students, it is unsurprising that research highlights many students frequently make ineffective revision choices and do not have sufficient knowledge around maximising memory retention. AI Assist solves this key problem using cutting-edge artificial intelligence, making revision simpler, quicker and helps students retain more.
AI Assist is a revision application that combines advanced yet established study techniques with machine learning to adapt to a student’s learning ability and optimize their revision, enabling them to make the most out of their revision time and long-term memory. AI Assist is a spinoff product from Verge, which started as an idea for the founder’s final year dissertation project and quickly became a great opportunity for the founder and Verge.
Starting with research, the app focused on solving areas like prioritization and memorization via established psychological research and revision techniques. Research quickly pointed to the benefits of spaced repetition for enhancing long term memory. With research also highlighting reinforcement learnings ability to adapt from multiple simple decisions over time the foundations for the app started to form.
From months of research and development a unique machine learning algorithm was developed and was tested on a dozen students. An experiment was used to evaluate the applications success, with the algorithm being enabled randomly for half the students. The results showed strong potential, highlighting that student who used the algorithm scored on average 52% higher results than the other students and retained 2.8 times more information when retested.
Currently, the application is in the process of being developed into a mobile and web app for deployment. Being developed in Flutter, AI Assist will have a single code base that can be deployed to iOS, Android and Web. Additionally, Flutter enables iterative development and feedback while minimizing development time and cost without sacrificing quality.
Overall, the work in this project intends to see AI Assist successfully implemented and ready to support students and to have a strong codebase ready for further research, development and advancement.
Contact us
Have any questions or want to find out more? Contact us, we love discussing anything related to technology, business or strategy.