India Tech Architecture Manifesto

Most citizens meet the state through a website before they meet it anywhere else. Too often that first meeting feels like walking into a badly lit office with a queue and no signboard. Broken flows, security holes, forms that only work on desktop. The message is obvious: nobody cared enough to fix this.

India now has a UX department. Good. But having one is not the same as using it well. Public sites should look and work like we expect our public spaces to look: maintained, readable, safe. Fast load times, mobile layouts that actually work, security that would pass a basic audit. That is not decoration. People trust institutions they can use.

A farmer checking a subsidy. A student downloading an admit card. A shop owner filing GST. These should not require a tutorial, a phone call to a helpline, or luck.

A lot of people running India's tech policy learned to build software before cloud was default, before mobile-first was obvious, and long before a two-person team could ship in a week with AI tooling. That is not an age joke. It is a gap problem.

Incubator boards, funding committees, and training programs should include people who live in today's stack. Not because young people are always right. Because the distance between how software gets built now and how it was taught twenty years ago is wide enough that decisions start drifting.

AI is already changing how products get designed, tested, and shipped. Treating it as a slide in a deck, or a weekend workshop, is how you end up with policy written for a world that stopped existing in 2019.

Hackathons are fine. They should not be the main way India finds its best builders. A weekend sprint and a trophy do not replace months of work on a hard problem. They give you a high, then send you back to a system that never cared about the work in the first place.

We need something permanent: a way for young builders to show what they have actually shipped and get credit for it. National recognition for people who fixed real problems at national scale. Logistics in districts with bad connectivity. Language tools for communities official software ignores. Systems that worked when the official ones did not.

Find these people where they already are: GitHub, public demos, social feeds, products people use without being asked. The point is not to run a contest. It is to spot builders early and give them a reason to keep going, including the audacity to build products that compete with Google, Apple, Microsoft, Amazon, and Meta on quality, not just price.

Every hour stuck in paperwork is an hour not spent building. Founders do not quit because they lack ideas. They quit because incorporation, GST, and the compliance stack around both still run at bureaucratic speed while the market runs at internet speed.

Company registration and GST enrollment should be fully online, trackable, and done within one business day for standard cases. Everywhere. Not as a pilot in two states with a press release.

Startup ideas do not wait in queue. If the first mile is slow, the product mile never starts.

Research is one of the highest-value industries on earth. In India it is still often treated like the job you take when the "good" private sector role did not work out. That attitude costs us. You cannot import your way out of every hard technical problem forever.

Innovation does not show up because someone said the word at a conference. It shows up when labs get funded, when researchers can make a living without leaving the country to be taken seriously, and when a smart kid can say "I want to do R&D" without getting laughed at.

Stop treating R&D like a footnote in the budget. It is what everything else in this document depends on: better gov software, stronger startups, products that can sit next to global incumbents.

India talks about wanting more startups, then taxes them at every stage before they have revenue, product-market fit, or a payroll. Angel tax debates, capital gains friction, compliance costs on day one. The message to founders and investors is the opposite of what the slogans say.

Early-stage startups and the people funding them should operate in a genuine tax-free zone for a defined window: no income tax on reinvested profits, no angel tax on qualifying investments, no capital gains on exits reinvested into new Indian startups within a set period.

This is not a permanent loophole for every company forever. It is a runway. Give ideas room to become businesses before the tax code treats them like Reliance.

Governments buy a lot of software. In India, too much of that money goes to legacy vendors and well-connected integrators while young startups building usable consumer products cannot get their first institutional contract.

Some of that procurement should go deliberately to apps and services built by Indian startups. That is not charity. It is how you signal to private markets and to founders watching from the sidelines that homegrown products are worth betting on.

When the state uses modern tools built here, startups get reference customers, real revenue, and proof they can handle daily use at scale. That matters more than another demo day.

Some consolidation is normal. What is not normal is one company owning an entire category with no serious alternative left. If Swiggy is the only food delivery app that works in your city and it goes down, everyone is stuck. Same story for telco stacks government runs on, or consultancies that keep buying every promising startup until competition is mostly cosmetic.

India is too large and too varied to run critical services through single-vendor choke points. Lock-in feels efficient until the one vendor fails.

Put hard restrictions on acquisitions that kill real competition in sectors tied to public infrastructure, essential services, or government procurement. The point is not to block growth. It is to keep options open for citizens, for startups, and for the state.