What 400 campuses taught us outside the metros

India has 1,338 universities. We work directly on 400 of them.

Getting to that number took eight years and it is the least transferable thing this company owns. You can buy advertising in a day. You cannot buy a relationship with a placement cell in a district town, because the only currency it accepts is having turned up before.

Here is what turning up taught us.

The gap is not talent

This is the finding that reorganised how we think about the entire market.

The distance between a graduate from a top-ranked engineering school and a graduate from a college nobody outside the state has heard of is real, and it is much smaller than hiring behaviour implies. On the assessments we run, under the same conditions, with the same marking, the distributions overlap enormously.

What differs is not capability. It is who shows up in March.

At the top of the ranking, recruiters arrive on campus, run a process and leave with an offer list. Placement is a scheduled event with a calendar and a queue. Two hundred kilometres away, at a college whose graduates test comparably, nobody comes. Not because those students are worse, but because the cost of getting there does not pencil out for a recruiter who can fill the same seats without the drive.

The bottleneck is distribution, and it always was.

Distance is not the variable people think it is

We assumed, early, that the useful split was metro against non-metro. It is not.

What actually predicts whether a campus works is whether one person there owns placements and has done for several years. That is it. A well-funded college with a rotating administrative post is harder to work with than an underfunded one where the same faculty member has run the careers desk since 2015 and knows exactly which of her students can do what.

Institutions do not have memory. People do. Every one of our 400 campuses is really a relationship with somebody, and the ones that lapse are almost always the ones where that person moved on.

Access is not the same as readiness

The first version of our campus work was too simple. Bring the listings to the students, remove the distance problem, and let capability do the rest.

It did not work, and the reason was uncomfortable. A student who has never met a recruiter does not know what a recruiter is listening for. They undersell real projects, they describe coursework in the language of coursework, and they have no idea which of the things they have built is the impressive one.

That is not a capability gap and it is not a confidence problem. It is an information gap, and it is the accumulated advantage of having been in a room where these conversations happen normally.

Credentials turned out to be the fix, for a reason we did not anticipate. A paid, proctored, human-reviewed result is a claim a student does not have to know how to make. The credential makes it for them, in a form a hiring manager already knows how to read. It removes the requirement to be good at self-presentation in order to be hired for something else entirely.

Why we still send people

The obvious question is why a company with an AI agent flies people to district towns to run hiring sessions.

Because the agent works better afterwards. Every session produces verified identity, real intent, and evidence of capability on the same person, and there is no other place in India where those three land together. That dataset is what makes the product good, and it does not exist unless somebody goes.

It is also the honest answer to what happens if a larger company decides India matters. They can build a better model than ours. They can outspend us on every channel. What they cannot do quickly is be the company that has been coming to a college in Jharkhand every February since 2019.

400 down. 938 to go.