The first time ViberNet went down during a shutdown order, I sat in the server room and thought about water. There is an old idea that the nature of a thing is defined by its relationship to what it lacks. A river is defined by its banks. A vessel by its emptiness. An ISP in Kashmir is defined by the government's ability to switch it off.
We had built around this, or tried to. OLTs distributed across the city. Multiple fiber routes into the core. Redundant peering where we could afford it. But a Section 144 order doesn't care about your topology. It reaches further than any cable you lay.
I started ViberNet because the internet here was bad. The more honest version: I started it because the internet here was controlled. These are related problems, but they are not the same problem. The first is an engineering problem. The second is something else entirely.
What Bad Infrastructure Actually Costs
People outside Kashmir understand internet shutdowns as a political issue. They are right, but they miss the texture of it. A shutdown is not just an absence. It is the meeting that doesn't happen, the invoice that doesn't get sent, the student who falls behind and can't catch up. It is the slow erosion of every economic activity that requires connectivity — which, increasingly, means every economic activity at all.
When we started laying fiber, the incumbents were offering copper-based DSL connections at speeds that would have been underwhelming in 2008. The pricing was high and the reliability was low. The customer service, if you could reach it, assumed you had no alternatives. You didn't.
The business case for ViberNet was straightforward: build better infrastructure, price it fairly, and the market is there. What I didn't fully understand at the start was how much of the business would be about building trust as much as building networks. In a market where people have been failed repeatedly, the pitch is not the speed test. The pitch is: we are from here, we are staying, and we will be here when something goes wrong.
Owning the Stack
We use MikroTik for routing — the hardware is affordable, flexible, and there is a deep ecosystem of expertise for it in this part of the world. FreeRADIUS for authentication. WireGuard for anything that needs to be encrypted end-to-end. The OLTs sit at distribution points across the city, and we run our own fiber to the last mile.
Every decision in that stack was a decision to own something rather than rent it. This is expensive in the short term. A managed solution would have been faster to deploy. But a managed solution also means that someone else's outage becomes your outage, someone else's pricing decision becomes your pricing decision, and someone else's relationship with the authorities becomes your problem at the worst possible moment.
In a place where you cannot always rely on external systems, you build your own. This is not stubbornness. It is the most practical thing you can do. — Something I understood only after building it
The lesson, which sounds simple and took time to fully learn: the more of your own stack you own, the more degrees of freedom you have. This is true for internet infrastructure. It is true for every other kind of business too.
What We Are Building Toward
ViberNet is not finished. We are expanding coverage. We are improving the reliability of the last mile in areas where the geography makes it difficult. We are thinking about what connectivity looks like in five years, when satellite options have matured and the economics of fiber have shifted again.
But the deeper project — the one I think about on the longer time horizon — is about what it means for Kashmir to own its own infrastructure. Not just internet, but the full stack of systems a modern economy depends on. Power, connectivity, payment rails, data. Each one that gets built locally is one less dependency on a system that might not have the valley's interests at heart.
I am twenty-one. I have time to work on this slowly. The problems are not going anywhere, and neither am I.