backing AI-native founders from first cheque to IPO.

Everywhere
meet boundless ventures. AI-native global seed capital. from india, for the world. boundlessvc.com
8
3
53
13,338
financial services in India is entering its most interesting decade. how money is lent, invested, insured and regulated is being rewritten. the institutions that built these markets and the founders reimagining them will shape what comes next, and they’ll get there faster together. @theboundlessvc and @OpenAI are bringing both to the same table. stay tuned for the unbundling.
1
3
8
296
we have a stellar lineup for our physical AI session on the 22nd of October on evals, retraining, and reality, co-hosted by @Qualcomm_Dev prateesh awasthi, co-founder at @armatrix: building snake-like robotic arms for hazardous, confined industrial spaces. @suvonilc, co-founder at manav robotics: automating high-precision, unglamorous work on automotive and component production lines. @levelheaded_94, co-founder at @fpv_labs: building infrastructure that turns human movement into training data for dexterous robots. @itsmanesh, co-founder at @Flo_Mobility: building autonomous robots at work on construction sites. they’ll talk us through what’s working in the field today, how they evaluate it, and what happens when things fail. physical AI will be won by whoever learns fastest from their failures in the field. link to register in the comments.
1
6
14
510
vertical specific manipulation dataset for specific industrial task don’t exist. the best way to build them is to deploy where the task is narrow and the outcome is clear. @armatrix_IN is building snakelike robotics arms for hazardous and constrained sites: one arm, configured per task, and every confined space adds to its intelligence. thank you to the @EconomicTimes for featuring our portfolio company. on 22 october, we're bringing founders and operators deploying physical AI in India together in bangalore, with @Qualcomm_Dev the @armatrix_IN team will be there. link to register in the comments. @puranchdET
2
3
14
643
almost none of the data from real robot deployments is making it back into training. the evals aren't standardized. the annotation pipelines aren't shared. the retraining infrastructure is being built privately. and staying there. that's the conversation we're not having publicly. evals, retraining, and reality: post-deployment physical AI. october 22nd, bangalore. co-hosted with @Qualcomm_Dev. everyone in the room is deploying. link in comments. @carrycooldude @localhosthq
1
4
10
518
boundless ventures retweeted
the most hyped model in robotics right now completed 7 out of 100 real manipulation tasks. taking a cap off a marker. passing a ruler from one hand to the other. getting a box open. the demo reel skips the other 93. if you're deploying physical AI, you already know this. the video gets the attention. the 93 are your actual job: the evals that catch them, the failure data you collect, the retraining loop that makes run 101 better than run 1. that's what this session is about. evals, retraining, and reality: post-deployment physical AI oct 22, bangalore. co-hosted by boundless and @Qualcomm a closed, technical room for founders and operators deploying physical AI in India right now. 15 seats. register now, link in the comments
2
3
15
1,440
the most hyped model in robotics right now completed 7 out of 100 real manipulation tasks. taking a cap off a marker. passing a ruler from one hand to the other. getting a box open. the demo reel skips the other 93. if you're deploying physical AI, you already know this. the video gets the attention. the 93 are your actual job: the evals that catch them, the failure data you collect, the retraining loop that makes run 101 better than run 1. that's what this session is about. evals, retraining, and reality: post-deployment physical AI oct 22, bangalore. co-hosted by boundless and @Qualcomm a closed, technical room for founders and operators deploying physical AI in India right now. 15 seats. register now, link in the comments
2
3
15
1,440
the saas seat was never really pricing the seat. it was pricing the best proxy anyone had for value more crm seats, more salespeople, more pipeline. AI breaks that proxy. when an agent processes 10,000 insurance claims on its own, there’s no seat to count. there’s a workflow. our founder, @natashamalpani, spoke to @ETEntrepreneur_ on where this actually lands. tokens and API calls measure what the system consumed, not what the customer received. the pricing unit should track the customer’s own value metric claims processed, hours saved, outcomes delivered. usage-based is the transition state. outcome-based is where this ends.
8
1
14
561
four teams. four themes. forty five minutes to build a globally competitive spacetech startup from scratch. the themes on the table: in-orbit servicing and debris removal, space manufacturing, orbital data centres, and re-launch infra. one team looked at all of them and decided to go their own way- battery packs for satellites. the moment the timer started, strategies diverged. one team picked a theme within 20 seconds, rushed to a conference room and shut the door to keep their trade secrets confidential. another spent the first ten minutes stress-testing every theme before committing. all of them got very technical, very fast. when it was time to pitch, all teams turned brutally honest and winning a fund vote from a room of people with deep expertise was not easy. the team that deliberately chose the theme they had the least conviction in, to push themselves, walked away with the most yes's. that was zero to orbit. more of these coming.
3
13
560
over the last few years, spacetech startups across launch, in-orbit and downstream have been going through an evolution. launch is reusability-first, because you want to make the most of each rocket. in-orbit is laying next-gen infrastructure for what comes next. and downstream is addressing faster communications and richer earth data. we’ve mapped the companies founded and funded since 2022. thank you to @livemint for covering the momentum this wave is generating, and featuring our portfolio company @piersightspace and our founder @natashamalpani. link to the article in the comments. @ghoshrwit @GangulyNabodita @SkyrootA @PixxelSpace @SanyarkSpace
2
4
23
791
india's quest is to build a global space economy. the talent and policy movement are here, can the capital keep up? our founder, @natashamalpani was featured by @ETtech on why our portfolio company @piersightspace is a global outlier. their data assets will only become more irreplaceable. on that note: if you're a space-tech engineer, we're hosting a space design workshop this friday in ashok nagar, bangalore. registration link for zero to orbit: a spacetech founder simulation is in the comments. article link also in comments. @SmruthiNadig
2
2
17
610
most spacetech engineers know their function better than almost anyone in the world. very few get to think across product, capital, team, and first customer at the same time under pressure, from scratch. zero to orbit reverses that for one evening. pods of 4-5 engineers and researchers pick a theme orbital servicing, space manufacturing, orbital data centres, relaunch infra and have 45 minutes to build a globally competitive startup. then they pitch. curated on the 18th of september in bangalore. link in comments.
1
12
518