Cuemath.

A child stares at a problem. Nobody rescues her.

It is a Tuesday evening, somewhere across one of 80-odd countries, and a nine-year-old is stuck on a fraction. Her tutor is right there on screen, watching every keystroke on a shared whiteboard. And the tutor says - nothing useful. A hint. A question. A cue. Then waits. This small, deliberate refusal to rescue is the entire business. Cuemath runs roughly 200,000 of these standoffs, and it has built a company on the belief that the pause is where learning actually happens.

Cuemath is an after-school, live, 1-on-1 online math tutoring service. Sessions run two to three times a week on its own platform, Cuemath Leap, built specifically so a single student and a single trained tutor can work a problem together in real time. It is not a video library. It is not a chatbot. It is a person, watching you think.

Most kids don't hate math. They fear it.

Before Cuemath, Manan Khurma spent years running a coaching outfit called Locus Education, drilling more than 10,000 teenagers for the IIT-JEE - one of the most punishing entrance exams on Earth. He was good at it. He was also, by his own account, troubled by what he kept seeing: by the time students reached high school, the cracks in their foundations had set like concrete. You could coach around the symptoms. You could not undo the years.

The diagnosis was unglamorous but stubborn. Math is taught as a pile of procedures to memorize and reproduce on demand. Reproduce enough of them and you pass. Understand none of them and you panic the moment a problem looks unfamiliar. The fear isn't of math. It's of being asked to think on your feet about something you only ever memorized.

So Khurma did the irritating thing successful coaches rarely do: he walked away from the lucrative end of the funnel - the exam-prep teenagers - and went after the eight-year-olds. The customers who can't yet spell "calculus," let alone fear it.

One man, a wager on the long game.

Cuemath launched in 2013 as a largely solo bet. Khurma grew up surrounded by scientists - a chemistry-professor father, a biologist mother - and had been teaching math to 17- and 18-year-olds while still an undergraduate at IIT-Delhi. The conviction was simple and slightly heretical for a tutoring business: teach math as reasoning, coined later as MathFit - the idea that mathematical strength, like physical fitness, is built rep by deliberate rep.

The model started on the ground, with home-based learning centers across India, before the bet got bolder: rebuild the whole thing as live, online, 1-on-1, and sell it to parents anywhere on the planet. Investors noticed. Sequoia Capital India (now Peak XV), Google, Alpha Wave and others put in a combined ~$124 million over the years. The pitch they bought was not "an app." It was a teaching method that happened to need software.

A whiteboard, a tutor, and a stubborn silence.

Everything runs on Cuemath Leap, the company's own live-class platform. At its center is MathCanvas, a collaborative digital whiteboard where student and tutor share the same workspace - audio, video, chat, and a curriculum aligned to the student's school and to US Common Core. Two to three sessions a week. Worksheets and assessments after each lesson. A separate app that lets parents watch progress reports roll in, which is either reassuring or terrifying depending on the week.

Big backers, a hard winter, and the numbers in between.

The case for Cuemath is partly the company it keeps. Google is an investor. So is Peak XV, formerly Sequoia Capital India. Alpha Wave led the last two big rounds. Time Magazine put it on its 2024 list of the world's top edtech companies. That is the flattering half of the ledger.

The honest half: edtech had a brutal couple of years, and Cuemath felt it. In 2023 the company cut roughly 100 jobs, then another 100, shrinking from a peak near 2,500 employees to a few hundred. Khurma came back as CEO. Revenue in FY24 came in roughly flat - around $14-15 million - though losses narrowed sharply. None of that fits neatly on a tote bag. All of it is the actual story.

Make the kid who loves math at 8 unafraid at 18.

Strip away the funding rounds and the layoff headlines, and the mission has barely moved since 2013: teach math as a way of thinking, not a set of tricks to survive the next test. The company calls the goal becoming "invincible in math" - which is marketing, sure, but it points at something real. A student who reasons her way to an answer owns it. A student who memorized it is renting.

That is also why the founder keeps coming back to AI. Khurma has said edtech is at an inflection point - capital is conservative, technology is shifting, and AI is rewriting what a tutor even is. For a company whose whole method is the deliberate pause before help arrives, the interesting question is whether a machine can learn to wait, too.

Back to the nine-year-old and her fraction.

Return to that Tuesday evening. The fraction is still there. The tutor is still saying almost nothing. But something has shifted in the few minutes we were away: she tries something, it's wrong, she tries again, and this time it holds. No one told her. She found it. Multiply that small private victory by 200,000 students across 80-plus countries, and you have the entire argument for why Cuemath exists.

The edtech gold rush is over. The companies that promised to replace teachers with screens have mostly learned a humbling lesson. Cuemath's wager was always quieter and, possibly, more durable: keep the human, keep the pause, and trust that a kid who learns to sit with a hard problem at eight becomes an adult who doesn't flinch at one. Whether AI sharpens that bet or undercuts it is the open question. The fraction, for now, is solved.