Devansh Swarup 👨‍✈️

428 posts

Devansh Swarup 👨‍✈️

Devansh Swarup 👨‍✈️

@swarup_devansh

Investments @ https://t.co/9dZIcVLmWl | Ex-Chief of Staff @ Lossfunk, Prev. Co-founder and CEO @hq_paperplane

Bangalore, India Beigetreten Mart 2020
2.3K Folgt367 Follower
Harveen Singh Chadha
Harveen Singh Chadha@HarveenChadha·
- elon will do anything to challenge the dominance of anthropic and openai in coding space - never bet against the power of having a strong post training stack (composer)
SpaceX@SpaceX

SpaceXAI and @cursor_ai are now working closely together to create the world’s best coding and knowledge work AI. The combination of Cursor’s leading product and distribution to expert software engineers with SpaceX’s million H100 equivalent Colossus training supercomputer will allow us to build the world’s most useful models. Cursor has also given SpaceX the right to acquire Cursor later this year for $60 billion or pay $10 billion for our work together.

English
6
3
93
3.3K
himanshu
himanshu@himanshustwts·
yes, chip / infra companies have always been early to IPO than any model company but its interesting that hottest model company today is merely 5 years into the market and going for the IPO soon. so if you are hot model comp / neolab / infra / app layer with good public pereception + effective distribution, the chance of going public will not be too far. (can be 2x faster than any!) watching out for databricks, mistral and cursor.
himanshu tweet media
TechCrunch@TechCrunch

AI chip startup Cerebras files for IPO techcrunch.com/2026/04/18/ai-…

English
2
4
60
4.6K
Neil Agarwal
Neil Agarwal@regalstreak·
YC startup school india wraps tomorrow we’re parking a full party bus outside the venue and taking the first people who get on for free drinks! 🍻 who's coming? VCs, founders and the coolest builders in bangalore! comment “PARTY BUS” for the invite + location. limited slots.
English
121
7
177
42.2K
Devansh Swarup 👨‍✈️ retweetet
Shubham Gupta
Shubham Gupta@shubg·
The best founders don’t just build products. They build communities around a shared obsession. @emergentlabs gets this. #vibecon2026 is proof. Across our portfolio, the companies that compound fastest are always the ones that make their users feel like they’re part of something bigger. Watching this play out live is a privilege. This is the place to be right now. 🙌 @scaletogether @mukundjha
Shubham Gupta tweet mediaShubham Gupta tweet media
English
1
2
11
522
Devansh Swarup 👨‍✈️
Devansh Swarup 👨‍✈️@swarup_devansh·
Super excited for this Hackathon, OpenClaw is slowly becoming the defacto "harness" to use AI Let's see what the ecosystem builds :)
English
0
0
2
156
Paras Chopra
Paras Chopra@paraschopra·
Optimism today is more needed than ever before. Given how many reality-distorting bubbles exist out there, pessimism is the easy choice. But those who believe that the future can be a happy place will indeed find ways to have a more satisfactory life than those who implode.
English
22
21
350
11.3K
Yogesh Singh
Yogesh Singh@_yogeshsingh·
feels like something incredible is starting to happen in #India 🇮🇳 i came all the way from Mumbai to attend multiple community events being hosted by @OpenAI in #NammaBengaluru #FOMO fascinating to talk to engineers from diverse backgrounds building on frontier tech #CodexBLR
English
1
3
38
1.6K
rakshith
rakshith@rakshith_twt·
hey @swarup_devansh, hope you’re doing well. I was really interested in attending the Community Dialogue with OpenAI by Together Fund, but I noticed that registrations have already closed. Is there any chance I could still join or be considered for attending?
English
1
0
0
51
Suhas Sumukh
Suhas Sumukh@suhasasumukh·
nah, small VCs aren't dying - they're just specializing harder. sub-$250M funds still do like 40-50% of seed deals and honestly outperform mega funds on returns bc they need way fewer wins. what's actually dead is small funds with no real edge or thesis what's printing. small funds world-class at something - deep in a vertical, own a geography, or have unique access to specific targets. blume/prime/stellaris in india, uncork/haystack in US - all small, all doing fine. large funds have deployment pressure that wrecks seed-stage decision making. small funds can take weird bets on unproven ideas. that's literally where returns come from.
English
4
1
40
1.1K
Yuvraj Aaditya
Yuvraj Aaditya@AadityaYuvraj·
Some conclusions. > VCs will get bigger. > Smaller ones will really need to get good at finding seeds. > All the seeds will be in accelerators/residency. > Landscape changing w solofounding - projects wont need as much VC help coming out residency. > Small VC ded unless you partner up
English
3
0
28
1.8K
Suhas Sumukh
Suhas Sumukh@suhasasumukh·
a moonshot worth pursuing: a palantir-style ai procurement system for the indian govt.
English
16
4
151
10.6K
Devansh Swarup 👨‍✈️
Devansh Swarup 👨‍✈️@swarup_devansh·
@vijayshekhar And then we say “We CaN bEaT ChInA” I understand this was shared in jest, but our pride for our “old ways” + false sense of confidence because of rich heritage caps the growth mindset
English
0
0
0
10
Vijay Shekhar Sharma
Vijay Shekhar Sharma@vijayshekhar·
Perfect autonomous ride and in India !
English
64
84
1K
91.4K
Devansh Swarup 👨‍✈️
Devansh Swarup 👨‍✈️@swarup_devansh·
@dhruvtrehan9 just automated AI research! He never formally studied CSE/Engineering, self taught himself how to code and built a system that can do ML research all in < 1 year 🙌🏻 “YOU CAN JUST DO THINGS”
Lossfunk@lossfunk

🚨 Releasing our technical report: Why LLMs Aren't Scientists Yet @dhruvtrehan9 tested if LLMs can perform end to end ML research. 3/4 attempts failed. One worked and led to a paper accepted at Agents4Science 2025, world’s first conference for AI authors. In the report we document six failure modes and four design principles. 🧵

English
0
0
3
91
Zikang Jiang
Zikang Jiang@ZikangJiang·
just moved from Sf to Bengaluru! where are all the hardware/robotics ppl?
English
232
73
2.7K
236.6K