Posted by 
Does it need MIT degree? Could India build next Cursor?
I can bet any above avg software engineering team in Indian GCCs knows how to integrate Developer interface with APIs like Cursor or handle scale. Indians have Sales capabilities too.
I worked in GCCs where we handled 900B high-value transactions monthly globally with team of 25 in 2018. Achal’s last project was rolled out in 140 countries, together on continet level data aggregation.
India definitely doesn’t have capability issue. What most pessimists portray or want to push to structural reforms.
Many cursor-like products (Not necessarily for coding) can be built.
Like, we at Vaayu built an AI layer on top of existing digital infra backed with a Deeptech AI infra for decisioning.
Then why it was Cursor? Not from India.
Before someone points to structural changes VC eco-system, University numbers etc etc. Let me share the exact simplicity that Cursor & its backers executed.
How an Unknown startup valued > Fortune 50 company & IT giants
I was working on BMW Autonomous driving in Germany…Complex AI systems. When news spread Tesla is setting up in Berlin….Entire Germany shook. Cz Elon will eat all automotive companies alive.
My first thought was – If this is Germany fearing US. wt about state of AI in India? This was around 2019-20.. ..pre ChatGPT era…Diff world
I was at that time, figuring out my next career paths. Observing Research, Business, Tech, Ideas.
I instantly knew – the opportunity is too huge. But it needed a different approach. Jobs, Indian IT can’t do this.
That’s when Achal & myself decided to come back from Germany & US respectively, to build Vaayu.
𝐒𝐢𝐦𝐩𝐥𝐢𝐜𝐢𝐭𝐲, 𝐒𝐩𝐞𝐞𝐝 & 𝐍𝐞𝐭𝐰𝐨𝐫𝐤
Simplicity – Simple is good, need not be out of the world.
I have been questioned why at Vaayu we build AI Native versions of Sales & finance tools.
Cursor built AI Native IDE that has been around for decades.
Especially VCs insist too much, too early on some experience that they have not done.
At Vaayu, we are clear one & majoritythe context is building what customer understands & is used to.
Cursor didn’t begin with a sophisticated cognitive architecture. It started with a simple, useful product. The hidden reasoning, retrieval and validation loops came later after they got fuel aka capital, a larger team, and resources.
𝐅𝐮𝐞𝐥 𝐟𝐨𝐫 𝐬𝐩𝐞𝐞𝐝:
𝐂𝐮𝐫𝐬𝐨𝐫 𝐫𝐚𝐢𝐬𝐞𝐝 $175m 𝐚𝐭 𝐯𝐞𝐫𝐲 𝐞𝐚𝐫𝐥𝐲 𝐬𝐭𝐚𝐠𝐞𝐬. Once things work a bit – This is a valuable asset.
You have some company structure – founders, a little brand, a few customers…it proves – there’s a market & possibility. This is where Vaayu – AI for Finance & Sales is today.
But, I am 100% certain, you can’t build viral GTM motions except a few lucky cases without fuel. More important is the compounding structure. How far you go with GTM is mostly dependent on the underlying fuel or cash flow.
𝐘𝐨𝐮 𝐧𝐞𝐞𝐝 𝐭𝐨 𝐠𝐨 𝐯𝐞𝐫𝐲 𝐟𝐚𝐬𝐭. 𝐁𝐫𝐞𝐚𝐤 𝐭𝐡𝐞 𝐭𝐡𝐢𝐧𝐠𝐬 𝐢𝐧 𝐒𝐭𝐚𝐫𝐭𝐮𝐩…𝐍𝐨𝐭 𝐬𝐚𝐧𝐢𝐭𝐲 𝐥𝐢𝐤𝐞 𝐭𝐨𝐤𝐞𝐧 𝐜𝐨𝐬𝐭𝐬, Unit economics or too much logic 𝐚𝐭 𝐭𝐡𝐢𝐬 𝐬𝐭𝐚𝐠𝐞.
Cursor & its investors put freaking 1600cr even at early stages.
Again, for comparison, Hardly 20-50 VCs in India may even overlap with the thesis of investing 5-10cr for similar tech bets. (My internal studies for our Seed round)
Cursor would be easily dismissed as a wrapper & not too differentiated
Did pedigree play a role?
Definitely yes. But to get the network, not the capability.
Some people get the network & timing right as God’s gift. Most others build it themselves.
There are 1000s of examples where pedigree seldom translate to big outocme.
But in my experience, High-agency is better bet than pedigrees.
Here’s how Cursor used capital aka fuel for speed
Cursor gave discounts…it costed them more to pay to Models than they charged from customers.
Increased usage to be top brand, top usability.
You can’t compete with Github & MSFT by being too rational….or thinking like 1000s of companies.
Below is Vaayu’s internal study done in March’2025, on how much it takes to support users. Of course, to grow at speed, you will have to sustain a burn. We don’t have Cursor’s number yet, but you get the gist of point we are making
(We tried to normalise the cost to DAU than what chatGPT published to show somewhat bigger numbers. chatGPT numbers aren’t OAI numbers; these are only for chatGPT)
Cost of major population-scale platforms as of March’2025
Average cost of talent needed in India
Engineering & Sales talent costs huge.
There’s physical limit to what founders or humans can achieve by themselves.
Greatest asset in startup is Company structure & team, validated direction….not the outcomes yet.
At Vaayu, we been at similar stage. Small, things work, positioned well & aiming to scale team.
An engineer will cost around 30-50L INR per annum, Doable in India
Given that you to build a serious engineering team, I am basing it on what I know about salaries in GCCs. To practically have a good engineer who can scale & build an engineering platform with 3-7 years experience will minimally need 30-50L INR on the conservative side to get them to startups.
That too, when supported by ace tech founder like Achal & myself.
But still peanuts, if you had to build it in valley.
Achal & I know such engineers in Investment banking, Autonomous driving & Data at continent scale from our jobs. But we know those engineers need clear instructions, vision, and motivation that only tech founders can provide.
𝐍𝐞𝐭𝐰𝐨𝐫𝐤, 𝐏𝐨𝐬𝐢𝐭𝐢𝐨𝐧𝐢𝐧𝐠 & 𝐓𝐢𝐦𝐢𝐧𝐠 > 𝐓𝐨𝐨 𝐦𝐮𝐜𝐡 𝐥𝐨𝐠𝐢𝐜
Cursor isn’t an engineering marvel. It’s not top-notch innovation like foundational models.
Compeltely doable in India.
Cursor’s success rests on a network & structure of resources, an engine & quick decisions by founders & investors with front loaded conviciton apart from capabilities.
India lags all except on capabilities.
Being at MIT gave them access to OpenAI & more in a circle who understand how you need to be more unconventional than too much logic in startups.
But network too can be built. What a waste of Indian’s god gift of being social & build a network will be, if not utilized for building big economic outcomes..
What it takes to build Cursor in India
𝐈𝐧𝐝𝐢𝐚 𝐝𝐨𝐞𝐬𝐧’𝐭 𝐥𝐚𝐜𝐤 𝐜𝐚𝐩𝐚𝐛𝐢𝐥𝐢𝐭𝐢𝐞𝐬….Our engineers & businesses are too smart.
But startups & even innovation are more about “𝐖𝐡𝐚𝐭 𝐈𝐟 𝐭𝐡𝐢𝐬 𝐰𝐨𝐫𝐤𝐬”….Not – “Why this would work”
Powered by Simplicity, Speed & Network
More about “What if it works” about Vaayu
Beyond Obvious – Decision Intelligence backed by Deeptech AI infra
AI Co-workers backed by Decision Intelligence, Applied AI reasoning