A Guru for every Indian
The Prime Minister has promised AI training for one crore youth. The scarcity that forced India to ration learning is ending.
By Rahul Patwardhan and Sriraj Mallick, CEO of Infopro Learning
On August 15, from the Red Fort, the Prime Minister promised AI training for one crore young Indians and a free online coaching network. That is a statement of vision: every young Indian, not a lucky few, equipped for the economy being built around them. The question that follows every vision is machinery. What do we build so the promise becomes ordinary reality?
Part of the answer appeared in The Hindu last week, in a lead essay proposing an AI tutor for every child, built as digital public infrastructure. The idea is right, and it can go one step further. A child in Class 6 matters, but India's most urgent learner is older. She is entering skilled work this year, with crores arriving behind her, in an economy hungry for productivity now, not in 2040. The promise is to teach one crore young Indians about AI. The larger opportunity sits one step beyond it: letting AI teach nearly thirty-seven crore young Indians everything else.
The momentum is real, and so is the distance still to cover. Only 4.2 percent of working-age Indians have received formal vocational training. The India Skills Report 2026 counts 44 of every 100 graduates as not yet employable, and Mercer-Mettl's Graduate Skill Index 2025, after assessing more than ten lakh students, puts employability at 42.6 percent. One in three young Indians, the International Labour Organisation finds, is neither working nor studying nor training. And the window is not open forever: the 15-29 cohort, nearly 37 crore today, will shrink to about 24.5 crore by 2036. Effort and money are not the missing pieces. The next push has to change the design itself, because a factory model that gives every learner the same course at the same pace can count certificates, but it cannot grow capability one person at a time.
The root of the design is scarcity. Good teaching was always scarce, so India built a system to manage the shortage: standard courses, large batches, and a handful of narrow gates through which everyone must pass at the same age. Selection had to stand in for development. A single score began to shape the next forty years, and the brand of an institution became lifelong proof of a person's worth. Every Indian family knows the weight of this; there is no need to argue it again here. What matters is the assumption underneath the whole design: that learning shaped to one person cannot be given to everyone.
That assumption has just failed, and India has known the alternative longer than anyone. In the guru-shishya tradition, one guru worked with one student at a time, shaping the path and the pace to what that student could best become. Modern research agrees: in 1984 the education researcher Benjamin Bloom measured that a student taught one-to-one performs better than 98 percent of students in a classroom. The guru was always the best way to learn. There were never enough gurus, so for centuries that advantage belonged to the few. AI now makes a guru buildable for every Indian: software that knows what you already know, sees where you are stuck, adjusts the pace, and stays with you until you can pass the skills test a real job role demands. When one-to-one teaching stops being scarce, learning no longer needs to be rationed. Standards do not fall in this world; the path to them widens. A readiness bar that anyone may attempt when prepared, the way a driving test works, can replace a ranking you get one shot at, at eighteen. Each person can grow at their own pace and have each skill validated when it is actually there. Skill becomes a unit that moves through the economy, alongside degrees and institute brands, not in place of them.
Nothing here asks existing institutions to step aside. The great campuses will keep their gates, and families who can reach them will keep reaching. This infrastructure is for everyone the gates were never built for: the marginalized, those without resources, the crores who live far from any serious skill-development institution. Their ability was never the problem. Access was. India's growth needs both doors open.
India knows how to build this kind of change. When the state wanted every citizen identified, it built Aadhaar, not a hundred thousand new offices. To move money, it built UPI, a shared standard that any bank or startup could plug into; private builders did the rest. The world now studies this as Digital Public Infrastructure. We built these rails for identity, payments, and documents. Learning never got its full layer, though the bricks exist: APAAR, the National Credit Framework, the Skill India platforms. Each is a strong start. Together they could be the foundation.
Call what India must build the Kaushal Chakra: the capability development loop, running on four public rails. At the base, the record: APAAR, the credit framework, and the skilling platforms unified into one system, with a skills passport the citizen owns as its face. Above it, the tutor: an open standard for AI tutoring, published as UPI's protocol was, that lets any university, startup, or ITI plug in. Then the human layer: the best teachers and institutions offering mentorship as add-on services wherever a learner needs more than software, so a great educator reaches thousands instead of forty, and is paid for it. Holding it all together, trust: an independent authority that continuously benchmarks tutor quality and certifies skill claims, so a passport entry means the same thing in Kochi and in Kanpur.
For the citizen, these rails add up to a single experience: a wheel that turns through three stages, each one the Indian tradition of learning would recognise.
Bodh is awareness. The record watches you learn over the years and shows you what you cannot easily see alone: what you already know, what you are close to, where your aptitude points. The marginalized and those without resources do not lack ability, and neither does the young person in a district with no serious training institution nearby. What they lack is someone who spots their talent and asks, have you considered this. Bodh is that voice for every citizen, and one principle must be written into it on day one: the system may build awareness and offer suggestions; the decision always stays with the person.
Abhyas is well-rounded development, not practice alone. The guru coaches you through what blocks you, teaches at your level and in your language, and gives you practice until the skill is real, with human mentors joining as an added layer wherever software is not enough. Because abhyas runs on your own time, it fits around a job or a family. Growth no longer waits for an admission cycle.
Siddhi is proof. When you are ready, you attempt the readiness bar a real job role demands. Whoever taught you never grades you; an independent layer validates the claim, and the proven skill is written into your passport, where an employer in any state can trust it. Siddhi is the unit that moves.
Then the wheel turns again. Each siddhi sharpens the next bodh, and capability compounds across a working life. That is the design: not a pipe that ends at a degree, but a wheel that never stops turning.
Three worries deserve straight answers. Access: Digital India has already done the hard half. Nearly 97 crore Indians are online, and a gigabyte of data that cost 269 rupees in 2014 now costs about eight. The connection reaches almost every hand; what rides on it must now include a teacher, on a cheap phone, in every Indian language. Teachers: nothing here replaces them; the services layer pays them to do what software cannot, which is to mentor and to inspire. Employers: no one is asked to abandon campus hiring. If the passport proves reliable, it becomes one more channel for finding people, and trust will grow at the speed of results.
Start where the Prime Minister has already pointed. Run the one crore AI-training programme on these rails, as the first cohort. Then form capability missions: groups that map the jobs emerging in India, from chip plants to green energy to the care economy, and build validated pathways into each. Judge the whole effort on two numbers: how much faster people learn, and whether they earn more a year later. Not certificates issued. Capability created.
The median Indian is not yet thirty. A developed India by 2047 will be carried by how much this generation learns, and how fast. India taught the world that money can move in seconds on public rails. The next proof is waiting: that human capability can move that way too, at each person's own pace, with no ration line in sight.
Comments