Helping founders and brilliant people with work visas, immigration, and starting businesses worldwide.

Joined April 2009
154 Photos and videos
Pinned Tweet
The US admits about 200,000 international students from India every year. 4 years later, they graduate into a free-for-all fight for green cards. Fewer than 3% will stay long term. The student visa-to-green-card pipeline is broken, and it hurts everyone. Quick math:
145
626
3,780
1,217,248
To celebrate the World Cup, I made a draft/tournament game you can play with your friends. It's totally free. Up to 12 people. Snake draft: pick 4-8 teams. 4 tiers of teams, from juggernauts to "just barely qualified" Underdogs get more points; the worse the team is ranked, they more they can score. Tech people - yes I used Fable. I also built in fun stuff: players to watch, team history, fun facts, betting odds... so you can pick teams to watch and what to pay attention to. Comment below and I'll DM a link for you to start a league.
1
1
250
Hosting a thing for international founders to understand visa options. If you're applying to YC, Speedrun or other accelerators, this is for you. International student looking to start something? This is for you. Drop a comment/ emoji/ whatever and I'll send you the details.
211
10
241
19,778
Almost 10 years of using Superhuman, but their AI responses can’t draft: “Let’s find a time to chat. Here’s my calendar” like I have used probably 10,000 times. What are smart people using for email these days? I want something that sounds like me and gathers context
2
4
1,234
kinda sick of the fraudsters might just fuck around and make a tool to help USCIS identify O-1 cases with fake media mills, pay-to-play journals, BS awards, judging events, etc. might make things tough for a ton of “profile builders” and the firms that enable them ¯\_(ツ)_/¯
5
61
19,127
Huge sigh of relief - just got MY O-1 visa approved. This was my 5th O-1 approval, and the weirdest, most complicated one yet. My wife jokes that I'm trying to set a record for Person With The Most O-1 Approvals. I'm definitely not. The stress of waiting for processing has probably shaved years off my life. But I've now hit just about every O-1 application type possible: 2018: New O-1 visa to join Legalpad, consular processing 2020: 3 year renewal/extension to stay at Legalpad (they CAN be done) 2022? Change of employer after Deel acquisition 2023: Asked to be put on a PIP at Deel so I could take a vacation and get paid / not get laid off :) Concurrent O-1 visa so I could quit, agent petitioner application submitted from Japan 2026: Agent petition amendment/ extension for Compass Visas GTM Ventures (consulting agent petitioner/ agent visa entity). Huge shout out to Kim, our attorney, for discovering and fixing all the issues an overpriced boutique law firm included in my previous petition, like my supposed employment as an Industrial Engineer 🤔
6
86
10,433
My take on Manifest raising $60m to be a law firm/ tech company… The question is: How does a company take a $750m valuation and get to a venture-scale return? At present, they’re maybe pulling $10m to $15m in services revenue with 100 lawyers on the platform. They have significant costs to advertising, acquiring customers, and tech development. At 3,000 cases handled, I’m guessing break-even to slightly profitable at best, probably underwater. Before the announcement, I wasn’t aware there was a tech angle to the company, I thought it was just a group of lawyers running under shared advertising. We know the angle isn’t selling tech to companies. They’ve correctly identified that lawyers are the worst tech purchasers on the planet. Props to them. So if it’s services revenue that leads the way to an $8 billion valuation to make VCs happy / interested, they need 4-8 billion in services revenue at a 1-2x multiple. (Maybe there is some hope of a 10-20x “tech” multiple, but when SAAS multiples are falling, the last thing I would expect is to see law firms multiples increasing) The US immigration market is not that big ($~8-10bn), and the story being told can’t be that they will eat up the entire market. I’m guessing the pitch is: “As we get bigger and acquire customers, our systems will get more efficient and drive up profitability per case. We can become an AI-native Big 3 law firm faster than Fragomen or BAL can make the switch, and we’ll eat their market through better outcomes and faster services.” They are probably also looking sideways, outside immigration, at other verticals: contract law, incorporation, acquisitions mergers, etc My mental model is currently a PE-style rollup of different attorneys/law firms, with some shared back office and tech that may make attorneys more efficient. Love to see new models and tech being explored in the space, this will be interesting to watch.
9
1,550
The full story is one of the craziest founder (and O-1 visa) stories I can remember. Not my story to tell, but Lumian will be big and Robin will share some learnings one day :)
Almost let down my closest friend at the worst possible moment… My friend Tauseef flew from Pakistan to San Francisco on June 1st, 2025. > one of my closest friends from Harvard Business School > an AI consultant who's been building in the space for years > one of the sharpest people I know A few weeks earlier, I called him and said, "Come. Let's build together." He packed up his life in Pakistan and got on a plane. On his first day of work, before he'd even opened his laptop… The venture I had brought him into fell apart. I was the one who had brought him to America. And I had just lost the ground under both of our feet. It got worse. Walking away from that venture meant walking away from my O-1 visa. Which meant I had 60 days to leave the country. Everything I had built toward in the US would be gone. I walked Tauseef outside and told him everything. I looked him in the eye and told him I was sorry and that he'd need to figure out what came next on his own. He looked straight back at me and said, "Robin… I came here to work with YOU. We'll figure it out. And you'll remember this moment." So I pulled out my phone and took a photo from that bench in Alamo Square Park where we were sitting. The next morning, Tauseef and I grabbed coffee before I flew back to Boston. I boarded a 6-hour flight that same night with 60 days on the clock and no idea what I was going to build. But the ENTIRE flight, one idea kept pulling me back in. Again. And again. And again. - I had built and sold a brand on Amazon - I had spent a year going deep on AI agents - I had watched agencies fail Amazon founders for years, including my own brand This was the most obvious thing I had ever seen. I landed, sat down at my kitchen counter, and opened my laptop. I already had a domain name from the AI newsletter I had been running for two years. The newsletter was called Lumian. I had been building that name for two years before I even knew what company it would become. 60 days to leave the country. One idea. A domain name I'd been sitting on for two years. That was the starting line. Something BIG is being announced tomorrow.
2
1
10
4,187
🚀 Second O1 visa approved — 15 days. One shot. Zero RFEs. This time it is to build my own company. Super late to post this (been 100% heads-down building my deep tech company for months), but this win is too big to stay quiet. O1 transfer petition is a massive filing as exhaustive as the original including all fresh reference letters. My first O1 visa? Approved in 3 days, also zero RFEs. Immigrant founders trying to build in America : US immigration is your number one priority. Not the product. Not the pitch deck. Immigration. I started shaping my extraordinary ability profile back in 2020 - five ears ago — before I even had a company. Treated it like a science experiment: studied every successful O1 petition, built my own protocol. Got questions? Drop them in the replies. I will try to answer every single one. And now onto building the Biomanufacturing powerhouse of America 🧬🚀💊.
25
4
90
9,637
Polsia is the AI equivalent of a podcast bro selling a “millionaire mindset” course. If AI could reliably launch 6,000 real businesses, why would they be selling a $49/mo subscription for people to generate their own businesses? If these businesses could generate an average of $25/ week, they’d be sitting on $7.8m ARR. But… they don’t. So they sell a subscription to people hoping they can be financially free by letting a computer do the work. Brutal to see smart people falling for it and lending legitimacy.
3
733
Ian doesn’t miss. If you’re an engineer looking to get in early somewhere…
One day soon, you’ll be able to turn a clever idea into a product over a weekend, then press a button and immediately turn that product into a fully operating business that makes money for you on its own. You’ll be able to instantly generate a fleet of agents that incorporate your business, open your bank account, create your website, sell and market the product, process payments, do your accounting, and file your taxes. At Synthetic we’re working 6 days a week to bring that day closer. Our first product is a completely autonomous AI agent for software companies, coupled with its own accounting system built from the ground up for agent use. We’re a team of startup veterans backed by Khosla ventures, and we’re looking for the next two product engineers to add to our team, in-person in San Francisco. If you want to work with an experienced team that’s serious about winning, consider joining us: synthetic.ai
4
3,701
yc founder applying for an o-1 visa: > no work history > no traction at startup >pre-product, revenue, everything > no press about work (because no work history) > $60k salary in sf (not highly paid) > no publishing or judging New immigration startup: “You’ve got a great O-1 case!” USCIS: “Denied. You can’t be extraordinary in your field if you don’t have any work experience in that field.” Immigration startup: “Unlucky! Let’s try again!”
24
7
560
78,067
Openclaw but it doesn’t spend your money or post on clawbook or anything, it just keeps you logged into all your saas stuff so you don’t have to fuck with 2FA and password managers all day.
1
7
1,531
Can’t convince me it’s not ChatGPT slop
1
729
Hot take? Prediction markets aren’t really that hot. Kalshi and Polymarket just used “prediction markets” as a back-door into legalized sports betting in California. The smart economics/politics/industry stuff has no volume. 1400x more money on NFL game that doesn’t matter
1
10
3,169
i get the gossip on cofounder breakups all. the. time. often, its a forced situation and founders never aligned on the plan is there data on cofounder breakups with “founders that added a cofounder for funding or accelerator” vs “ cofounders building together pre <event>” ???
13 Dec 2025
that’s all that matters. thanks for this, yc. fyi: this was 25 days ago, before the decision deadline.
9
3,255
the entire culture war over immigration right now comes down to one group that thinks they want a meritocracy and free markets while also begging for protection from competition
1
13
1,938
It’s bad for aspiring immigrants but watching an immi company pay their way into an O-1 / EB-1 community and then absolutely fumble by 1) not knowing their audience 2) using lawyertalk that ppl don’t understand 3) giving incorrect advice to simple questions kinda hilarious.
3
31
5,864