Let me take you back to 2021 when I met the just-launched Rocket Learning team. The model was simple. They leveraged WhatsApp, a tool that was already on everyone’s phone, to help Indian children get school-ready. In two years time, they served 1.75 million children. This week, they won Google.org AI for the Global Goals. Their charge? Use AI to improve early education for 30 million Indian preschoolers in the next five years. In this exclusive chat, meet two of the brains behind the big vision: Rocket Learning co-founders Namya Mahajan and Vishal Sunil. We dig into the impact of this big win, what’s next on the AI roadmap, and more.
Shannon Farley: Namya and Vishal, first off, congratulations on winning Google.org AI for the Global Goals! It is a big accomplishment. Can you tell us about the AI challenge and the problem you’re setting out to solve with AI?
Namya Mahajan: Sure! As Google knows better than almost anyone, AI can be a game-changer for solving the significant challenges the world is facing. Google.org AI for the Global Goals is one of the ways that Google is demonstrating its commitment to harnessing AI for the greater good. At Rocket Learning, the tech nonprofit I co-founded, we share the belief that tech can meaningfully solve entrenched social problems and so, our focus is on pre-literacy and pre-numeracy skills for low-income Indian children.
Picture this: almost half of children in an Indian classroom can’t read a single letter by the time they reach second grade. We believe that reaching these children as early as possible is key to breaking this cycle. So, we’ve set up 200,000 WhatsApp groups, where a staggering 1.75 million children with their parents get to dive into the exciting world of patterns, stories, numbers, and play, all before they turn six years old. They’re supercharging their reading skills, improving their ability to understand pictures, and so much more.
The Rocket Learning platform is super user-friendly, so it works well on basic phones with apps like WhatsApp and YouTube. And our team is truly passionate about leveraging cutting-edge technology like AI to support communities that need it the most. Google.org AI for the Global Goals is paving the way for us to do more of this, and at scale. We’re honored to win alongside an exceptional group of co-grantees, all of which are deploying innovative solutions for incredibly complex problems.
Farley: What does this win mean for Rocket Learning and the millions of children you serve?
Vishal Sunil: Winning this challenge is a significant achievement…not just for Rocket Learning, but for the families we support in low-resource areas. It means we can use AI to enhance children’s learning experiences and empower parents and community educators to support learning at home.
Our win could potentially benefit millions of people and improve education for future generations. Google.org’s support brings Rocket Learning one step closer to turning our long-standing dream into reality: seeing every child school and life-ready so they can escape poverty and reach their full potential.
Farley: It’s clear that AI has the potential to help tech nonprofits solve some of society’s thorniest problems and unlock impact in ways we haven’t seen before. Namya, can you share more about how you anticipate AI will positively impact the families you serve?
Mahajan: We’re not just setting our sights high…we’re aiming for the stars!
As I mentioned, we’ve shaped early childhood education for 1.75 million preschool children since 2020. So far, the results of our efforts have been nothing short of amazing. Our users perform on par with the top 30% of their peers. Plus, parents are doubling their teaching time, building learning as a daily habit. As we look to what’s next, we’re looking to scale. To reach more young people, we’re going to work with Anganwadi workers, India’s childcare providers. The Indian government sponsors childcare and nutrition centers for its poorest communities. We are providing these childcare heroes with the essential tools and training needed to transform them into bona fide early education experts. We’ll leverage AI as a powerful force for social good — preparing more kids for school, supercharging graduation rates, and lighting the path to brighter futures.
Farley: Rocket Learning is using AI in unique ways – and from what I’ve heard, you’ve developed a fairly complex product. Vishal, you lead product at Rocket Leaning. Can you walk us through exactly how your team is leveraging AI and this Google award to achieve your mission?
Sunil: My favorite topic! One of our most impactful uses of AI is a system that automatically grades homework. It’s awesome. It’s been a tough build technically…we’ve been toiling away for 14 months fine-tuning one algorithm that can give pixel-specific feedback but is also generalizable to over 900 unique worksheets. It’s been hard! But we’re so excited about it.
We’re talking top-notch technology, like a convolutional neural network (CNN). If you’re unfamiliar with CNN, think about it as a highly specialized computer program that’s fantastic at understanding and analyzing pictures. When we apply it to grading homework, it examines the students’ answers just as one might look at photos. It’s really good at recognizing things like handwriting and numbers, making it a powerful tool for grading homework effectively and efficiently. It’s like having a super-smart assistant to help us assess homework assignments.
We’re also kicking it up a notch with Google’s advanced pix2seq model. What this means for the children we serve is that they’re getting feedback that’s super detailed, like a GPS for mistakes, helping them course correct quickly and easily.
But we’re just scratching the surface! Thanks to our collaboration with Google, we’re gearing up to make our system even smarter. It’s not just about grading anymore…it’s about creating new, tailored learning materials and experiences for each student.
Farley: Now that you have additional resources, what will you build next with AI?
Sunil: Thanks to this support from Google, we’re diving headfirst into the realm of AI. First and foremost, our AI system is set to fundamentally change homework grading. Instead of tedious, manual grading for teachers, the product will serve as a “homework grading assistant,” making grading incredibly efficient.
Also, we’re harnessing AI to make educational materials more engaging and personalized. By analyzing how our users interact with our platform and delving into the inner workings behind the scenes, we will provide learning experiences tailored to each child. Think of receiving content and lessons perfectly aligned with your interests and learning requirements, making the entire learning process more effective and more enjoyable. In a nutshell, our mission focuses on seamless integration of AI into every facet of our educational framework.
Farley: Google.org is a pioneering force in funding ‘AI for good.’ They’ve been funding ‘AI for good’ for nearly a decade. We need more funders like Google.org to step up to the plate. If they don’t, vulnerable populations may miss out on the benefits of emergent technologies like AI. What would you say to potential funders who are considering joining this effort to harness ‘AI for good?’
Mahajan: When funders invest in ‘AI for good,’ they’re unleashing its capacity for making a deep and real impact on society. Investing in AI for good fundamentally changes life trajectories for our most vulnerable. There are great ideas everywhere. Supporting these ideas can actualize creative solutions to make the world a kinder and more flourishing place for all.
Rocket Learning started three years ago and will be helping 2 million kids by next year. This was only possible because of technology and support building out the technology. But that’s not enough — our goal is to help 30 million kids in the next five years. This is a big challenge, but AI makes it possible.
Farley: Thanks so much for the work you do, Namya and Vishal!
Read the full article here
