Why the businesses software forgot are the best market in AI

If you asked a family business in India five years ago whether they would commission software built around how they actually work, most would have laughed you out of the room. Not because they didn't see the value. They did. But because the math was broken.

Real software was either an enterprise-vendor sentence or a dev-shop gamble, where the quote came in at ₹40 lakh and the thing that got delivered was a Bootstrap template with a login page. You either couldn't afford it, or you could and regretted it. So you did what every rational operator did: you bought SaaS, bent your business around it, and lived with the friction.

That math changed in the last eighteen months. And nobody has really said it out loud yet.

What actually happened

AI didn't automate the software industry away. It collapsed the cost of building good software for the problems most businesses actually have. Specifically:

• A small team running a serious AI pipeline can now carry the whole software side of a business that could never have justified an in-house one. That is a change in who can be served, not a change in how hard the work is.
• The part that historically ate the margin, keeping software documented and maintainable so that the next person can safely change it, got much cheaper to do properly rather than to skip.
• What did not change is the hard part. Working out what a specific business actually needs, before anything gets written, still takes the same conversations it always did. Building software has got faster. Building the right software has not.

So the interesting question isn't "what can AI do for huge enterprises?" Huge enterprises already had their software. The interesting question is: who was locked out before, and can they have it now?

The answer, for us, is: pretty much everyone we talked to growing up.

The businesses we actually want to work with

India has this enormous bench of service businesses and family-run firms (agencies, clinics, professional practices, manufacturers, distributors, traders) that have compounded wealth over decades. Real businesses with real customers and real operations. The kind where three generations of the same family have been building the thing.

Almost none of them had software built around how they work. The ones who tried either paid a multinational too much for too little, or got burned by a dev shop that couldn't ship. So they ran their operations on WhatsApp groups, Excel files, Tally, and whatever vertical SaaS was cheapest.

That's the space. Not greenfield startups. Not dev-first companies. Operators with real businesses and real wealth who never had an affordable path to software, automations, and AI that actually fit how they work.

And the same pattern holds globally. Every market has a middle layer of businesses too small for enterprise vendors and too specific for template SaaS. Until now, they had no good option. Now they do.

What that looks like with us

Not a ₹40 lakh "digital transformation." A few honest ways in, scoped against whatever is breaking your week:

• A partnership. For creators and operators with a real audience and no product behind it: we design, build, and run the thing your audience pays for, and take our fee as a share of the revenue it makes. A partnership, not an invoice. One or two a year. Read the terms.
• A fractional CTO. We become the software partner your business does not have, on a flat monthly retainer: decide what is worth building, build it, run it, and keep you current as the tools change. The sites that convert, the automations that quietly take hours off your team, the AI tools that give you leverage. An ongoing partner who ships the right things, not a one-off project that goes stale. Read the pitch.
• A one-time build. An experiential site or an interactive pitch deck, scoped and priced upfront, yours to own. The kind of thing that used to need an agency and a quarter.

We also fund, build and run our own software in the lab. Creator Kamai is live and you can open it, which is a more useful thing to judge us on than anything we could write here.

A paid discovery call to work out where to start, then we ship.

How it is priced

One price per tier, the same wherever you are. We do not price by how much we think a market will bear, and we do not price by the hour, because neither of those tells you what you are actually getting. What you are buying is one accountable partner who owns the software side and stays on to run what gets built.

But also: because the only way this works at scale is if we stay honest about fit. We take on small problems, big ones, and everything in between; we just pace each to the time it genuinely needs, and if you want a read on that, we are glad to give you an estimate up front. What we screen for is not the size of the work. It is whether you are a business we can do our best work for. We are straight about what we are good at and what we are not, and we would rather be the right fit than make the sale.

The thing we are actually betting on

That most of the interesting software in the next decade will be built not for Fortune 500s, but for the businesses underneath them: the ones with real cash flow and real problems who were priced out of software that fits for the last twenty years. That the people who win this market will be small, close to the work, and honest about scope. That AI is the multiplier that finally makes the math work for both sides.

If you run one of those businesses, you now have an option you didn't have two years ago. That's the thesis. The work is just what it looks like to execute on it.

• Puro