What I saw
- What I understood in the first five minutes
- Where exactly I got confused
- The point I'd have closed the tab
- One positioning line I'd test instead
2026
India has spent thirty years supplying the world's engineers. The question now is whether it starts owning the companies.
This is a close look at the people trying to answer that. One company at a time, read properly, with the founder's reply printed next to mine.
Read · Rebuttal · Repeat
Episode 01 · Applications open
The deal
The whole arrangement, in two numbers.
You give
45min
One call and a login. That is the entire ask, and there isn't a second one later.
I give
6hrs
I use the product the way a first-time customer would, read your site as a stranger, and map everyone working nearby. Then I write down exactly what I saw.
Five pages, within 72 hours. You get the last one.
Who else is out there
- Eight to twelve companies, mapped side by side
- What each claims, who they sell to, how they price
- Where you're genuinely different
- Where you sound like everyone else
Who actually buys this
- The buyer I think writes the cheque
- Whether your messaging points at them
- Two channels I'd test first
- Three investor questions you can't answer yet
What I'd do next
- Three things, ranked
- Each one doable inside thirty days
- What I'd want to see from you in six months
What you said back
- Your reply, printed in your own words
- Every place you think I've got it wrong
- And what I got wrong, once I had the real story
No pitch afterwards. There is nothing for sale.
Then it becomes an episode.
First
You read the whole thing before anybody else does. Nothing goes out until you've seen it.
Then
You send your reply and I print it next to mine, unedited, including the parts where you think I've got it wrong.
And
Cut anything you'd rather wasn't in there. No explanation needed. Or keep the whole thing private and we'll never speak of it again.
Most founders never get a proper piece written about their company. This is one, and you have the final say on every word of it.
A sample
What one of these actually reads like.
Episode 01 · Page 1 · What I sawExample
The first five minutes
Your homepage says you're an orchestration layer for production agents. After reading it twice I still couldn't tell whether you replace the framework someone already uses or sit on top of it. I guessed on top. Your docs confirmed it four pages down, which is three pages too late.
Where I'd have left
Signup asks for a workspace name before showing me anything working. I don't know what a workspace is in your product yet, so I'm making a decision before I have a reason to care. This is the highest-cost moment in your funnel and it's forty seconds in.
What I'd test instead
Something closer to what your own changelog already says in plain language: keep your agents running when the model changes underneath them. Concrete, names a pain your buyer has felt, and doesn't need the word orchestration.
This is a made-up example, not a real client. I haven't delivered any reads yet, so you'd be among the first fifteen. When there are real ones and the founders are happy for me to share, those go here instead.
Why
I'm at the same stage you are.
My whole working life has been inside the thing this is about. I dropped out of engineering in 2022 to build instead. Since then I've helped build a developer community past 300,000 people, helped put together a conference that reached 5,000, and I'm working on a company of my own right now.
So I know how it goes. In India a company becomes real the day it raises, and everything before that you do alone. You write your own homepage. You guess at pricing. You explain the thing forty times and still can't tell whether it lands, because everyone who hears it is either being kind or being polite, and those are different problems.
What you want is one person with no stake in it to sit down, use the thing properly, and tell you what they actually saw. I've wanted that and not had it. So I'm giving it away fifteen times, and doing the six hours rather than pretending a coffee chat is help.
The selfish part, out loud: I want to be an investor one day. Reading companies early is the job. I'd rather learn it by being useful to fifteen people than by having opinions from a distance.
And the country has enough people writing about the companies that already worked. Somebody should be paying attention while it's still uncertain.
Who
Early enough that an outside view still changes something.
Pre-seed to Series A
Far enough along that there's a real product to look at.
Building with AI
Any category. Infrastructure, tooling, applications, all of it.
India, or Indian founders
Based here, or building from here for somewhere else.
Apply
Five questions. Two minutes. No deck required.
I need this one to find your product.
Even one sentence helps.
Reads come out as episodes, and you approve yours before it does. Opens your mail app with the answers filled in. If nothing happens, copy them to hello@thetalentstate.in.