Three weeks in Bangalore: observations on India's AI startup ecosystem

Introduction

10,000 people who will really decide the direction of the next wave of AI. There will be multiple claims amongst 7 million developers but it will really be 10,000 who will decide.

That’s what Dr Santanu of MIT said to me in an VC tech event. My 3 week trip got me thinking about what is happening in the Tech capital of Asia and what was happening on the grassroots. Today I want to condense all of my learnings into three simple points:

• What is happening in the Indian startup ecosystem? Who is likely to go global?
• Who are interesting Ecosystem builders? What type of founders and people should go to them?
• Future predictions on where the Indian tech ecosystem

💡 Reality Check: These are my takeaways from 3 weeks of conversations in Bangalore (8th August 2025). Startup ecosystems move fast, opinions evolve, and I’m always learning. Drop me a message if you’ve got different insights or want to debate any of these observations.

So let’s get into it:

What is happening in the Indian start-up ecosystem? Who is likely to go global?

Based on my interactions in AI conferences, alot of indian startups are being built for what they call “Tier 2” and “Tier 3” cities. The companies attemping to go global are “Tier 1” + strong focus on expanding globally. These are generally the guys getting a VC paycheck.

Sarvam AI

One startup that I could not interact much with but looks promising is Sarvam AI. They are trying to build a whole swiss knife for full stack generative platform and are backed by India’s big companies and companies namely TATA, AADHAR, Infosys and even the government of India.

Notable mentions

NimbleEdge

I spoke to Neeraj, one of the founders and a genuinely interesting person who is building on-device AI. This field is something I am still trying to understand but it looks like a wild west field rn where people are trying to figure out where the foundational parts can go. This is definitely a deep tech company which could takes decades to realize but my interactions with Neeraj, a seasoned startup founder who worked in both the US and India tells me that they will make good judgement irrespective of how the startup goes. Check out their podcast

SarvMAI

I met Harish, one of the cofounders. They are building voice based e-commerce for nano business i.e for the third world model. He was an interesting person and here are my reasons why I found them interesting:

• founder with a deeply personal story. Someone who genuinely felt empathy for the average indian peddler or seller from a village.
• along with their technical cofounder, they built their VoiceERP in 12 Indian languages and 500+ dialects build in just $300K with 90%+ accuracy.
• discussing term sheets to expand in Indonesia, Thailand and even Eastern Europe.

What are the big VCs backing?

• Nexus Ventures are backing traditional businesses. Indian ecosystem despite the IT/startup boom is still heavily profitable for Consumer Tech + traditional Business.
• Accel is trying to push for precision manufacturing. Read here

Who are interesting Ecosystem builders? What type of founders and people should go to them?

I had a great 2+ weeks stay at Draper Startup House, Karan and his team are truly doing some great work here! Great team assembled and personalized connection to the average Indian start-up founder. They regularly have events including some great badminton sessions. Definitely a place to checkout. They have a young crowd that is really trying. It’s not the talent that impresses, it’s the energy and drive of people that I love!

I haven’t gone to e-chai events but they look like an interesting place.

💡 One observation I made is that Ecosystem builders, VCs and global companies are relatively siloed. This should be considered when investors and startups build for India.

South Park commons is a unique start-up accelerator incubator. They have an interesting -1 to 0 approach where they let people with interesting track record sit with other interesting people, have chats and just explore an idea that you care for 6 months. This is one of the best approach for deep tech. Definely chat with Ananya

Press enter or click to view image in full size

If you are curious to explore a technical idea but not sure whether you want to make it into a business |

💡Dr Ramesh and Santanu are engaging with the US + Indian ecosystem for their project Nanda. A push for standardized protocols when the world has agentic web. A place definitely worth checking out.

Personal observations to note

• Bangalore could be 3rd best startup ecosystem after London 2nd and San Francisco 1st. Based on my conversations with VCs and startups there, the dropoff from 1 to 3 is huge though.
• However the above statement is based on conversations and ground feelings rather than actual data. The actual data shows Bangalore at 14th in the Global Startup Ecosystem Report and 6th in global tech talent hubs according to Colliers’ Global Tech Markets: Top Talent Locations 2025 report.
• Groq looks like they are going to win the inference infrastructure market. With a heavy scaling push in India backed by Lightspeed and no Indian champion to fight the inference battle.
• There are some good principles being documented for startup founders to think about when scaling. Some can include network effects, proper validation, developing the right traits as founders to scale and getting the right sharks to back you. Perhaps an article I can write about this in future.
• Alot of AI startups are valued at 33 times their valuation. This makes them moonshot projects and only a few players are making money right now.
• With alot of Gen AI start-ups trying to build up hype, they would be better suited to capitalize on building and selling these start-ups as small to medium sized enterprise rather than raising a lot of money.

Future predictions on where the Indian tech ecosystem

• my predictions are simple:
• India needs to hit a home run in the next 5 years or VC market could dry up. A story similar to postman is needed for global investors to be bullish on India.
• India’s $1B+ deep tech alliance signals strong potential for AI innovation, though the government’s track record, from gaming app bans to UPI’s competitive advantage over private players, shows how quickly market dynamics can shift. This partnership will need careful navigation to balance ambitious policy goals with supporting private enterprise.
• Indian Deep Tech ecosystem is still minuscule compared to the US. Check this thread
• Europe is not even in the conversation for AI startups compared to what I saw in India. Only exception being London but UK isn’t really a part of the EU.