An open letter to India’s aspiring cloud entrepreneurs, and to the MSMEs ready to reclaim control of their data.
On 12 November, millions of people opened their laptops and found that their email, chat, files and shared documents had simply gone. For roughly two working days, a large part of the world’s office work stopped.
There was no cyberattack. The cause was a problem with settings applied to the provider’s own servers.
Think about what that means. An organisation can have hundreds of talented engineers, sophisticated monitoring, premium enterprise subscriptions and a carefully documented business continuity plan. And when the provider’s service fails, it can do exactly one thing.
Wait.
This was not the first time.
- A configuration update. In June 2025, an automated change at a major cloud platform cascaded into hours of downtime for a content delivery network and the consumer apps that depend on it.
- A geopolitical decision. In 2025, a major Indian company lost access to its own cloud tools overnight. Not a technical failure, resolved only after legal escalation.
- A DNS failure. In October 2025, a single region of another major provider took down banking, gaming and social apps around the world for most of a day.
- A settings problem. In November 2026, a third hyperscaler’s productivity suite went dark worldwide.
Four incidents in seventeen months. None of them was an attack. Three providers were undone by their own internal processes, at a scale where one mistake becomes the whole world’s problem at once. In the other, an Indian company’s data became hostage to a decision that had nothing to do with it.
Either way, the customer could only wait.
This is not a criticism of any provider. Even the best infrastructure fails.
But it forces an uncomfortable question:
Have we exchanged the complexity of managing our own infrastructure for the helplessness of depending entirely on someone else’s?
We have discussed this problem for years. In conference halls, in boardrooms, in articles like this one.
The time for discussion is over.
The time has come to build our own clouds.
Not One Giant. Thousands of Clouds.
When we talk about alternatives to hyperscalers, we imagine enormous data centres, billions of dollars and thousands of engineers. We imagine having to build another AWS, Azure or Google Cloud.
We don’t need to.
I have argued that enterprises are buying F-35s when the future belongs to drones. The same is true of cloud. A single giant is powerful, expensive and, when it fails, it fails everywhere at once. Many small, capable and coordinated units are resilient by design.
India does not need one Indian hyperscaler to challenge the giants. It needs hundreds, eventually thousands, of independent cloud service providers.
- A young entrepreneur in Ranchi, Patna, Mysuru or Coimbatore, serving local businesses from a few servers.
- An existing IT service provider adding secure email, storage, backup and hosting for the customers it already knows.
- A group of engineering graduates building a regional cloud business.
- And on the other side, thousands of MSMEs choosing a provider they can see, call and trust, over a distant hyperscaler.
The technology already exists, and most of it is open source: Linux, Proxmox, OpenNebula, OpenStack, Ceph, Nextcloud, Postfix and Dovecot. You can start with a rented server or a single colocated machine, a modest budget and a focused set of services, and grow as your customers grow.
These providers will not win by out-spending hyperscalers. They will win on what hyperscalers cannot offer: proximity, personal support, specialisation, transparency and accountability to customers they actually know.
Why We Cannot Wait
It is tempting to treat this as a problem for later. It is not.
Every year we wait, leaving gets harder. Each new workload, each terabyte of data, each workflow built around one provider’s tools adds another link to the chain. Lock-in does not stay still. It compounds.
AI will deepen the dependence. The next generation of business tools will run where the data already sits. If our data sits with a handful of global giants, so will our intelligence, our automation and our decisions.
And the window is open now. Mature open-source platforms, affordable hardware and skilled engineers in every Indian city: the ingredients have never been more available than they are today.
Today, building our own clouds is a choice.
Tomorrow, it may be a cost we cannot afford.
You Can Do This
If you have ever thought about starting a hosting or cloud business, and then talked yourself out of it, this part is for you.
- You don’t need a data centre. A rented server or a single colocated machine is enough to begin.
- You don’t need to know everything. Nobody does. Every infrastructure engineer you admire learned by building, breaking and fixing things, one problem at a time.
- You don’t need customers across the world. Your first customers are probably businesses you already know: in your city, your community, your network.
- You don’t need to do it alone. That is the whole point of this letter.
What you do need is curiosity, discipline and a deep sense of responsibility for someone else’s data. Everything else can be learned.
I asked earlier whether we have exchanged complexity for helplessness. The answer is not to go back to the complexity. It is to take back the responsibility, and tame the complexity. That is what I am working on: taking the hard parts every new provider struggles with and making them simpler, repeatable and well documented. Your energy should go into your customers, not into reinventing infrastructure.
And you will not be alone on this journey. I would be glad to support you with simple toolkits, practical guidance and mentoring, so that you can succeed, from your first server to a service your customers can rely on.
What I Learned the Hard Way
I say this with humility, because I have made this mistake myself.
Over two decades, I have built infrastructure ventures on open source, including security delivered as a service long before the term became fashionable. The ideas were sound. The technology worked. Customers were happy.
What held them back was never the idea, the market or the technology.
It was the same mistake every time. I tried to be the whole solution myself.
What I lacked was not capital. It was collaboration: other capable people, already in motion, whose effort compounds with yours instead of waiting on it.
That is why I am writing this letter. Not to build something for myself, but to give the next generation of infrastructure entrepreneurs what I did not have: experienced people beside them from day one.
You should not have to learn this lesson the hard way, as I did.
What I Hope to Offer
This is a personal initiative, pursued in my own time, with no commercial interest.
No consultancy fees. No equity. No commercial strings.
1. Free mentoring. I have spent more than three decades building and running infrastructure, from Linux and networking to email, security and the cloud, and I have made my fair share of mistakes along the way. I would like those lessons to save you time, and I hope to walk alongside you as you choose technologies, design infrastructure, understand costs and prepare to operate reliably.
2. The Open Cloud Entrepreneur Toolkit. I am burning the midnight oil on it right now, and I shall not sleep in peace until the first version is in your hands: open reference architectures, automation, and cost and capacity planning resources, so you do not have to work out every component from scratch. Start with whatever matches your market and budget today:
- One server: ISPConfig for web hosting and email, or Proxmox for virtual servers.
- Three or four servers: OpenNebula for multi-tenant infrastructure.
- Ten or more servers: OpenStack for a full self-service cloud.
It covers what installation scripts never do: planning, security, monitoring, backups with real restore testing and incident response, with simple checklists of what good looks like for hardware, network, security and operations. It stays open, with no obligation to use any vendor or service.
If you run an MSME, I want you to have a real choice beyond a distant hyperscaler. Work with an independent Indian provider who keeps your data in India and whom you can actually call when something goes wrong. Or run your own private cloud, in your server room or on colocated servers, built with the same toolkit and with a local partner to support you. What matters is not which path you take. It is that the choice is yours.
Let Me Be Honest
Running a reliable cloud business is not the same as installing software. Security, redundancy, connectivity, backup, monitoring, customer support and financial discipline are not optional. Some of you will find it harder than you expected.
The business opportunity will be yours to find: your market, your customers, your city. That part only you can do, and it is what will make your business truly yours.
But on the technical side, you will have a friend in your corner: someone who has learnt from more than three decades of experience, and more than a few mistakes, and who would be glad to help you avoid them.
An Invitation
If you are an engineering student, an IT professional, a small hosting provider, or anyone with the ambition to build a cloud business, I would like to hear from you.
If you run an IT services company, whether in hardware rental, AMC, networking or system integration, I would especially like to hear from you. Your customers already trust you. Cloud services may be your natural next step.
And if you are an MSME ready to reclaim control of your data, I would like to hear from you too.
You do not need to have all the answers. You do need the willingness to learn, the discipline to operate responsibly and the determination to build something that lasts.
I hope to share the first version of the toolkit, with reference architectures and practical guides, soon. After that, I hope to grow this into knowledge-sharing sessions, workshops and a community of entrepreneurs who learn from one another. Reach out now, and tell me what you need most. It will shape what I publish first.
Digital sovereignty does not mean every organisation must own its infrastructure. It means having real choices: control over critical dependencies, and the ability to recover or move when circumstances demand it.
The outages will happen again. Another configuration change will cascade. Another decision, made somewhere else, will reach into an Indian business that had no say in it.
The only question is whether, next time, we will be able to do anything more than wait.
We have waited long enough.
The cloud does not have to belong only to the giants.
The time has come to build our own clouds.
Uttam Jaiswal
If this resonates with you, connect with me on LinkedIn and send me a message, or follow this blog. I will share updates here first.
