Why we co-created Shunya with Teachers

Ekta Grover
November 10, 2020
10 Minutes

Repeat after me. K12 Learning is not a content problem.

“Be amusing: never tell unkind stories; above all, never tell long ones.”
-- Benjamin Disraeli

The prequel - behind every story is a Story (We know what we are doing!)

Shunya was born decoupled from COVID and yet ended up evolving so much because of it. There could not have been a better time to build a Technology of the future than this, because COVID was natural catalyst for the Infrastructure of the future of learning. In that it pushed our thinking multiple decades to build something of value.

Before we even started actively talking to Teachers for informative interviews, we talked to over 25+ parents, in tier 2/3 cities and a theme kept emerging. Content alone won't cut it, as the social realties of the learning spaces there are very different. Right from shortage, or temporal/ availability of the teachers, with an incessant complaint over overall quality of teachers. The parents were more and more concerned than they have ever been traditionally, partly fuelled by the Overall promise of the growth landscape in India and the opportunities across professional fields.

With targeted Fear-of-missing-out that most Edtech companies latch on the "Sharma ji ke bete ne to.." - all parents are looking at the mythical herd that doesn't exist.

There had to be a positive way to build something of long term value and to take our kids into the future of learning, with Technology. So, we turned over to where kids spend most of their learning time, in their schools. Not glossed over to some apps claiming to inceltivize the kids via fancy celebrity shirts, and virtual badges. Effectively producing outcomes that did not relate

After this, we spoke to 60+ teachers PAN India over 4 phases of interviews and it was clear that the most fulfilling part for a teacher is to see the students excel. Intent , personal responsibility & action, trying a new generation of teaching within the limits of the classroom, was not a problem at all. Yet inspite of raising the bar on personal attention in the classroom, teachers struggle with poor levels of parental confidence, classroom learning experiences that do not scale.   This was very counter intuitive to what the current EdTech solutions proclaimed - that the quality of teachers is bad. What we found was that the quality of teachers did have impact on overall long term outcomes, but those effects were not because of the teacher's alone.

It was a complex phenomenon happening in real time. And one that could be measured and orchestrated.

learning = f(reinforcement, micro concept, classmates, cognitive ability, parental involvement, environment stimulus, feedback, ° engagement  ….)

...But first back to the reality

Teachers are managing their classrooms shouldering more & more responsibilities, while they are spending about 40% time on teaching related non core activities.

40% now hold that number in your head! And these are tasks that could be easily "train" our machines to do it.

So, we started with a simple hypothesis - that effective deep learning is a personal experience, and effective learning should be a logical extension of our natural life and mental models. And if we could use Technology to assist Teacher to better scale learning outcomes, it would be a journey worth building.

So, Scaling better learning outcomes with Technology became our focal point, keeping Teacher & learner centricity at the core of everything we build.

First principles. Distinctive learning for distinctive learners

We know that every child learns differently. Which means that what they consume , how they consume the byte sized learning per concept is a hyper personal choice, tailored to their unique cognitive ability, socio environmental factors, parental involvement and comprehension ability. Students no longer have a linear exposure to concepts that aid classroom learning. Students learn from an array of exposure - text notes, visual concepts as images, audio, video, active learning from assignments. All of this means that the stimulus to a child is increasingly multimodal, too.

Consider this, when a child fails to solve a 2 digit mathematical problem, say (77-29), he or she isn’t failing in math but the underlying concepts leading up to it. Much like an involved parent would, what if we had a technology which could help diagnose this and predict the learning challenges with also helping the student to improve . And this we know by the most extensive Research ever done, by ASER, India , as part of which 546,527 children, in age group 3 to 16 years were surveyed . Many things stand out from ASER reports year on year - most important of which is that the gap widens with every year of academics. And the sooner we predict, detect, personalise and intervene - more is the possibility of long term meaningful difference to overall learning outcomes and raising confident kids who can self regulate their own learning.

ASER, 2018

This means there are micro concepts that make up higher order concepts that collectively make up what we call "knowing". And that this knowing much like any pattern, can be reverse engineered, within the bounds of current Technology.

The dropouts are not gradual, but they happen over the years. A phenomenon that can be predicted , and acted upon, with the right stimulus. Much like medicine, life altering conditions if diagnosed timely, can alter the course of life. And in learning specifically, highlight special gifts that neither the child, nor the parent is aware of .

But how do we know what stimulus works best for a child, and what if we personalise the stimulus to improve learning outcomes ? To give better long term learning outcomes, not just marks. So, what if we could measure the effect of this exposure in a child safe environment and improve the learning collaboratively?

Predict. Recommend. Improve, obsessively.

Much like the use of most advanced technology in Sports performance or the similar technology used to help us discover and consume content to our uniquely hyper personal entertainment choices(Netflix), Shunya is building the same technology and algorithms that powers similar hyper personal objectives - backed by most nuanced mathematics and machine learning. It is set in the context of a child’s own school - which means that it has access to the most high fidelity data about a child’s learning context. We model this through the most nuanced algorithms and mathematics to power learning back at schools .  

So, where do the current Edtech solutions fail short ?

The current EdTech solutions only have a 20-30 minute small fragmented window to a child’s world, which is insufficient for any meaningful diagnosis, a 360 view and intervention to support a child’s distinctively personal learning needs. They are largely content centric. Since the content relevance changes hyper locally, they fall short of any long term cognitive gains.

Interesting many experiments have been run to see the effect of short term calming music to exam / high stress situation performance, but the overall results were only temporal. This is exactly how short term radio button tests work. Since online radio button tests work on recognition and not recall, and children learn with familiar mental models, the learning outcomes in these platforms do not replicate at school, leaving the parents disoriented.

The problem is not the radio button format, the problem is doing it online alone, while involves low sensory  inputs and decreases the reliance on long term muscle memory which enables more long term connections.

The familiar mental models in K-12 learning are handwritten first, and fit with intuitive human centric communication modality. Which is why we created Shunya as a human centric learning platform.  

Learning will always be a community first activity. And we can predict a child’s ability and academic excellence based on in class exposure much more than any EdTech offering can.  This is why we built Shunya to overlay the best of technology on classroom workflows and injecting intelligent intervention, recommendation plans, predictive student regression  - a true leap in the way we have looked at learning as life impacting and life shaping activity.   

... And you will see why learning will NEVER be the same again.