đĄ âShould we make ice cream or chips?â asks Mr. Goyal. But hereâs the untold story of why weâre forced to make ice cream.
- Layak Singh
- Apr 5
- 3 min read

At a recent investment event in Delhi, Mr. Piyush Goyal made a powerful statement:
âShould we make ice cream (delivery apps) or make chips (semiconductors)?ââDukaandari hi karna hai?ââAre you proud we are creating delivery boys and girls?â
Itâs a bold and necessary wake-up call.But for many founders like me, whoâve been in the trenches â the reality is far more complex.
We Tried to Build the âChipsâ
When we were building Artivatic.ai, we werenât chasing quick revenue or valuations.We were building a full-scale AI Operating System for businesses â using vision intelligence, speech tech, decision AI â something that could fundamentally change how insurance, healthcare, and finance worked in India and globally.
We put our hearts and every penny into it.We mortgaged our house.We worked nights, weekends, and holidays.We survived without salaries for months â because we believed in the vision.
It took us 2+ years just to build the core product.We werenât delivering ice cream. We were building the kitchen, the recipe, and the machinery from scratch.
And Then, We Went to Raise Funds
We knocked on every door â investors, government bodies, banks.
But hereâs what we heard:
âWhere is the revenue in year 1?â
âAre you profitable?â
âShow us PAT in your P&L.â
âCan you give personal collateral?â
Institutions like SIDBI, Startup India, and other grant programs were full of processes and paperwork.They offered âš5â10 lakh grants â which is nothing when you're building IP-heavy deep-tech platforms.
No government-backed working capital.No low-interest debt.No real safety net.
And thatâs when the harshest truth hit us:
In India, bold innovation is celebrated on stage â but unsupported in the system.
So, We Pivoted
We had to shift to revenue-first models.We moved into productized services, started building platforms with faster go-to-market â not because we gave up, but because we had salaries to pay.People who work with us have families, children, EMIs, and dreams of their own.
Building deep tech is a luxury in India â unless youâre lucky or have runway support.
Talk is Cheap, Survival is Not
Starting a company in India is not like starting one in Silicon Valley:
It takes months to get approvals
You face a license maze across sectors
Thereâs no tax buffer for early-stage risk-takers
Collaterals are demanded even for innovation grants
So, no â we donât want to make ice cream.
Weâre forced to, because the kitchen for making chips doesnât exist yet.
The Real Questions We Must Ask Ourselves
If India has the largest number of STEM graduates, why arenât we building the next NVIDIA or OpenAI?
Why are we measuring innovation with short-term revenue metrics?
Why donât we fund patience and experimentation, the way countries like China, Israel, and the US do?
Why are entrepreneurs penalized for taking risks, and celebrated only after they become profitable?
To Build Chips, We Must Build the Ecosystem
Itâs time we stop pretending inspiration is enough.We need policies, debt, grants, equity capital, and government procurement programs that back early-stage deep tech.We need to fund missions, not just metrics.We need a system that gives builders oxygen in their most difficult phase â not demand results before theyâve had a chance to breathe.
Conclusion:
India wonât become a deep-tech superpower by telling founders to âthink bigâ â it will happen when we help them survive long enough to build big.
To all fellow founders out there struggling in silence:Youâre not alone. Youâre not wrong. The system just isnât built for you yet. But maybe, together, we can change it.
Read more: https://medium.com/@lsvimal/should-we-make-ice-cream-or-chips-a-harsh-truth-about-building-deep-tech-in-india-9a32c903bbb6
Comments