The problem Old Skool Analytics solves

At this point, all of us have been through online classes, courses and even exams. Despite the advantage of having a recording that you can watch anytime and the flexibility it provides, online classes are still not favoured by colleges and companies because of lack of monitoring. Since schools and colleges house thousands of people under one roof, they are going to remain closed for much longer than other industries and hence online classes are here to stay. Even though the classroom education has its own pitfalls, but one undisputed advantage is the ability of the instructor to continuously monitor the class and find out if some students are not paying attention. This monitoring is complex in nature as it is not necessary that if a student is looking at the instructor it implies that he/she is paying attention, as humans we can easily identify by someone's body language if they are in fact not paying attention, but this task becomes difficult in an online setting.
To overcome this, we have studied multiple research publications to find out what exactly implies attention. We read through the research of psychologists with expertise in this domain and have condensed it to create an algorithm, that detects and quantifies the attention of students by analysing multiple features including their eye movements, changes in their facial expressions and their voice. This AI-powered algorithm continuously monitors all three of these features through the webcam feed of the students and grades their attention LIVE.
The teachers can then review the attention score of each student through a live dashboard and talk directly to students with lower attention scores, to make the online classes much more interactive. The same software can be deployed by ed-tech companies to monitor students takin pre-recorded classes and ensure that the certificate of completion is only issued to students with a certain attention score, thus making these certifications much more credible.

Challenges we ran into

We faced multiple issues while building the project the first was to ensure that students shouldn't be able to tamper with the results by simply staring at the screen and getting high attention scores, so we reviewed lot's of literature in psychology to come up with an algorithm to monitor multiple features so the system isn't reliant on just one of them.
Another challenge was to make the dashboard that can show students' score live to the instructor, but we eventually overcame this by creating a portal wherein the dashboard is integrated.