Learning Analytics in Class Saathi

The education sector has seen a lot of reformation in recent years. It is not unusual since it is one field where progressive values are incorporated and included without much doubt or hesitation. The education sector survived the pandemic and all the various turbulences it presented for this very reason. It is one area where change is welcomed and technological integration is crucial in welcoming these changes as well. One such instrument that is being very widely researched and incorporated into education these days is “Learning Analytics”. 

Let us understand what Learning Analytics is and how Class Saathi incorporates it into its system.

Photo by Mikhail Nilov from Pexels

Learning analytics is a mechanism that collects vital data to assess key areas of learning and reports on it to drive impactful reformations. The logical framework that governs learning analytics is nothing new. It has been around for decades. So now why is it gaining so much novel attention? The answer is simple, learning analytics makes use of artificial intelligence that uses data smartly in order to not just report findings but also suggest measures that can accurately predict future outcomes based on these numbers. This hastens the process of data collection, interpretation and analysis and also gives educators tangible outcomes to work on so that change can be driven in the most efficient and minutest ways.

This takes us to the next part of this article – what are the learning analytic tools that Class Saathi incorporates into its system?


Class Saathi prides itself on being an app that gives personalised learning recommendations for every student that uses this app. This may sound like a tall claim but it is one that is rightly justified by the system that our developers in South Korea have worked so hard to build. For the purpose of this report, we spoke to the people who creatively built this system and this is what they had to say:

Personalised learning is at the heart of Class Saathi’s operations and we built an app that benefits every student who uses it. For this purpose, our developers built a system with three main components:

A. Knowledge Level
Every response is recorded and analysed by a system based on areas such as difficulty level and skills. The app first records the comfortable learning level of the student and gives them questions based on the recorded level and pushes them to gradually move up.

B. Updated Difficulty Level
The difficulty level for each skill or question is assigned based on the number of students who get it right in the first attempt. If there are more students who get one question or skill right, then the question is deemed easy and difficult if it is vice versa. This keeps changing with improved student learning outcomes.

C. Solve Interval Data
This is the most revolutionary data tool that we use. What this does is record questions that a student gets wrong consistently and recommend quizzes to help them learn the concepts attached to these questions. The system does this for a while and then gives the student the same question to see if they have learnt the concept and moved up the levels of difficulty. This prepares students to reach higher order levels or learning without making it a difficult process.

As you can see, the logical framework that governs the app uses very complicated metrics and building that can only have been the product of genius minds. It is no easy task to standardise a product for students with globally varying learning levels. 

Having said all of this, we must also remember that learning analytics can not be 100% accurate because it is constantly evolving. It can however predict the most probable outcomes for maximum benefit and maximum reformation.

In the words of our developers, it is a great difficulty to match reality to the learning level that is recorded by the system because accurate reading can be interrupted by many variables. However, the idea is to perfect a system by recognising the gaps, much as we do for our students. The important thing is to keep striving for perfection so that these gaps will be completely eliminated from our system and we have outcomes that don’t get in the way of achieving a higher understanding of student behaviour and learning.

Class Saathi’s IP-registered AI capabilities

Photo by Tara Winstead from Pexels

What does AI stand for?

AI stands for artificial intelligence. In this domain, intelligence is an artificial experience demonstrated by computers. TechTarget describes AI as “a simulation of human intelligence by machines”.

What does AI stand to achieve?

The agenda behind enlisting AI is simple. It is used as a replacement for human intelligence to run tasks, especially big in volume, that can diagnose problems, make decisions, solve challenges, predict situations and drive favourable outcomes etc, all using user-generated data.

Class Saathi’s AI-Powered Technology:

If you have been following our journey, you already know that Class Saathi is a digital learning platform that solves a multitude of problems emerging from the education sector with the help of Artificial Intelligence. Since the keyword Artificial Intelligence is used rather generously by tech companies around the world, there is a misconception that artificial intelligence used across various domains are all the same.

To help us establish some distinction, we spoke to AI developers at TagHive Inc., to understand how the AI technology they use for Class Saathi stands apart from their competitors. These are our findings.


Firstly, we need to understand AI in the simplest terms. Some human problems cannot be addressed by human solutions. It could be a problem that has many dynamic facets or the volume of engagement may be too high for humans to compute and process it fully. In these events, AI is used to capture the problem and drive the successive solutions. While machines and computers are doing this activity, it is the human brain that programs artificial intelligence to function accordingly. For this reason, AI has a lot of potentials to surmount many plaguing challenges across the globe in a variety of fields.

Having set the aforementioned tone, it is now imperative that we understand how Class Saathi deploys its artificial intelligence in the field of education.

Robot eating chips with a spoon from a bowl
Photo by Tara Winstead from Pexels

We started by identifying the challenges that were often left out from the problem-solving domain simply because of its challenging natures. We then built frameworks that would address all these challenges with one perfect solution. With time, it was observed that our solution could address so many more. This governing thought brought us the blueprint of our operation:

  1. Lack of accountability emerged as a general problem in the education sector with large gaps of communication between the various stakeholders, namely parents, teachers and administrators.
  2. Every student learns differently and at different paces. Catering to dynamic learning needs is a challenge that supersedes normal classroom instruction.
  3. Due to the diversity in learning needs across the education sector, targeting learning lags is a huge challenge for educators.
  4. Tracking student learning progress is often a challenge to address manually since it is not easy to conduct assessments frequently. The logistics involved are expensive and time consuming. Moreover, standard assessments are unreliable to determine progress since it is designed for a specific, often small demographic of students.
  5. Public education systems around the world, often the most populated in the education sector, see a high number of drop-out rates.

Listing out these problems shows you how diverse they are in nature and it seems humanly impossible to address them all with one solution. However, the AI developers at TagHive put on their creative thinking hats to create a system that not only tackles all these problems but also builds on them. Let’s see how:

  1. By creating a single app for all stakeholders, the Class Saathi developers created a transparent system of communication through which reports are shared, analysed and configured very fluidly. What it also does is use AI diagnostic tools to key in on specific focus areas of the problem and offer it to the stakeholders for tracking and monitoring.
  2. Class Saathi’s AI technology uses knowledge level diagnostics to accurately describe each child’s learning pace and level instantly. To top that, the personalised learning engine generates tests optimised for each level specifically catering to every student’s learning needs.
  3. The Rapid Diagnostic technology that Class Saathi’s AI enables allows teachers to see where exactly each student’s learning lags are. This is otherwise impossible to do manually, even in a small class of 50 students.
  4. Class Saathi allows for teachers to conduct assessments as and when they want to with no additional planning or cost. It saves time and money. What it also does more specifically is use knowledge tracing AI technology to analyse student reports based on assessments to predict potential progress for each student, based on their unique learning levels.
  5. Lastly, Class Saathi’s AI developers are planning to tackle student attrition through an AI capability called study pattern prediction that can warn educators when a student is at a high risk of dropping out of the system

Class Saathi occupies a unique position in the education system. We are a solution that acts as a personalised learning platform, a digital assessment solution, a diagnostic tool, a communication network based on evidence and so much more. Each and every aforementioned element is made possible by cutting edge Artificial Intelligence technology curated by some of the most creative minds in the industry.

Want to know more about us? Write to us at [email protected] to schedule a demo!