Getting ready to schlep, from project to corp
Part 1 of The Indian Deep-tech Guide for Ambitious Founders (IDGAF)
Note -
I hope for this to be an on-going series sharing learnings on building a deep-tech, deeply-ambitious startup in India. We’re calling it IDGAF because acronyms that mean something tend to be better than those that don’t and because it’s 10.40pm and I can’t think of anything better. And perhaps because you do really need to not give too much neural real-estate to what people think if you’re going to go after a moonshot.
I don’t claim to have any reason to be in the position to give useful advice except for genuinely going all-in for the past 2.5 years on building a deep-tech, deeply-ambitious company. We’re nowhere close to being successful but dognosis has grown out of their puppy stage, all gangly and starting to chew up shoes, having made genuine progress in our goal to build olfaction AI and solve for early cancer detection, and we continue to remain on the scent. Hopefully there’s something here that will turn out to be useful. Expect this to remain a WIP and feedback and comments always appreciated.
One of my favorite Paul Graham essays is Schlep Blindness. PG talks about how being a founder is filled with schleps, an originally Yiddish word for a tedious, unpleasant task. Having the ability to continuously knock out schleps can be a superpower because it allows you to see and access possibility spaces that others simply don’t because they couldn't be bothered with all the schlepping.
All founders have to deal with a fair amount of schlepping but the founder building in India does enough of it to warrant being awarded a whole degree. And it starts right from the very beginning - with incorporation.
“Wdym I just open a Stripe Atlas acct and whip up a Delaware C-corp, it’s easy”
Sweet summer child before realizing the glorious truth of the holy LLP-Corp-PvtLtd trinity
Before beginning the mega-schlep task of incorporating entities to do business in India, it’s useful to consider if you even need to do so. Sam Altman has a good piece - Projects and Companies - where he talks about why you should try to be a project for as long as possible, instead of becoming all official with a Inc or Ltd. Companies are ~serious, they are these mostly-fictional, at least in the short-term, entities that have to care about compliance and optics and having a sense of importance. You know, the opposite of hacking around and building things.
I was able to keep Dognosis, initially dogluk, as a project/idea/scheme for a solid 2.5yrs, before we incorporated in Delaware in late 2023. We did this to have something up by the time we raised money, one of the most common reasons to incorporate. Until then, I introduced myself as a cognitive science researcher interested in canine olfaction and wearable BCIs for dogs. This was great. I was able to spin up visiting research collaborations with labs in Israel, Hungary and the UK, expand my surface area without unnecessary constraints, and really focus on understanding how to go about the R&D of Dognosis, instead of wasting time talking to VCs, networking with ecosystem builders, or any of the various time-sinks its easy to fall into at that early-stage.
But at some point, whether to raise funds, pay yourself or team members, embark on a serious commercial partnership, or get your first contract - you will need legal entities to do legal entity things. And unfortunately, if you’re in India, the maze of decisions is far from simple.
Everything I’m about to say comes with the primary caveat that none of this should be taken literally (though I hope to write it such that it can be taken seriously). As always, take this as one input into your decision-calculus with the knowledge that I’m a first-time edge-case founder building a rather weird deeptech company. And of course, it’s definitely not legal advice.
Anyways. The first decision branch is - will your parent entity be based in India or abroad?
This is a big decision you’ll need to make, ideally at the very beginning, between incorporating only in India, with your parent entity as a Pvt Ltd, or to incorporate a parent entity abroad - such as a Delaware C-corp. Frustratingly, you need to make this decision in advance of truly having data on the real cruxes - where will you be able to most easily raise funds + where will most of your revenue come from.
For a deeptech company, where you will be able to raise funds is likely the main crux. Deeptech R&D is not cheap and real revenue tends to not be in the picture in the beginning, so fundraising is essential. Grants and prizes are amazing and should absolutely be actively pursued but the big ones tend to require a certain level of development, which require funding, a classic chicken-and-egg. VC capital tends to be the only real way to break out of this.
This means that choosing your parent entity can come down to optimizing for raising your first rounds of capital. In 2024, this is most likely a Delaware C-corp.1
To be honest, I’m not well-versed in the history of why Delaware . It’s a tiny state of 1M people, east of DC, that most founders eventually discover at this stage, for the curiosity that is their C-corp. I believe the general consensus is that this is downstream of Delaware case law on venture capital and startup being well-established and investor-friendly, and once this got set, you had a positive feedback loop making it the norm. As a founder, you want your product to be unique and stand-out but you really don’t want your entity structure to be weird and give investors additional friction, so going with the well-established tends to be good advice.
Perhaps the biggest boost of the Delaware C-corp, unlike the Indian Pvt Ltd, is allowing for your first rounds to be raised on YC standard SAFEs. This can be insanely easy in terms of pre and post-raise logistics. A founder and investor needs to agree on only two things - the valuation cap and the check size. Once done, you can generate a SAFE on Mercury or Carta in a minute or less, send a link to the investor and they can sign and wire in a minute and that’s it. The best investor interaction we’ve had was when an angel committed to a check on our first <30 min phone call on Monday, signed and wired the money on Tues, and the money hit our bank on Thursday. Investing in an Indian entity would have involved a lot more friction.
A decent number of investors will probably still put in the work (and tbh I don’t know exactly what this really entails) to invest in an Indian entity - ultimately, the right investor should be willing to back you even if it requires a bit of paperwork. But at the very early stages, where you may really only need 1-2 investors to bite but the number of firms willing to bet on you is already low, introducing friction could narrow the distribution, i.e. decrease your odds for luck, even further.
The Delaware C-corp is especially salient for the deep-tech founder in India because it is notoriously difficult to get funded by Indian VCs, especially if you don’t tick the boxes of a) PhD from elite University, b) significant traction such as multiple patents, pilots, etc, c) gray hair.
Ok cool. So the next step is to go on Stripe Atlas and incorporate a C-corp right?
If only.
If you’re an Indian resident and are planning to incorporate a C-corp (or already did), you cannot (legally) incorporate using an online tool (or at least not any online tool). Instead you mostly likely must surrender to this wonderful structure -

Continued in part 2, it’s too late to get into the wonders of the trinity right now and I’d rather face BLR traffic to get back home. We just moved our office to Malleshwaram!
The only exceptions are perhaps in space and semicon where govt support is not just valuable but a pre-requisite for success.
Useful article for founders
this is so cool!!