Any successful EB-1 applicants without PhD in Tech
Hey, Indian here in 100+ year backlog. Considering applying for EB-1 GC/NIW and trying to figure out a good route. No PhD, 5 papers in applied ML conferences top tier in their respective field, 100 < citations < 200, 5 patents issued and another 7 applied for, reviewed for top tier 5-6 conferences in applied ML and some workshops at top tier ML conferences. L6 Applied Scientist at Amazon. TC 420K. May be able to get recommendation letters from some profs I've met at conferences / taken on students from as interns (although many are increasingly ex-profs) and from an AMZN VP who gets decent media coverage. Since EB-1 isn't current for Indians my plan is to wait for a year or two for citation count to rack up and wait for some more patents to 'vest' then apply for EB1. Questions: 1) What would be a minimum 'safe' number of citations? 2) Portraying achievements the right way seems to matter a lot and Amazon lawyers seem conservative - what is a good law firm to use to hype me up to USCIS? 3) These days I'm getting a bit disillusioned with novel research and realizing drawn more towards management and owning customer-facing systems - would becoming a manager in the near-future harm my chances of getting the EB1? 4) Anything else I can do in the meanwhile to make myself look pretty for the committee? 5) Any other tips and tricks? (e.g. applying for EB-2 NIW, applying for extraordinary ability vs outstanding researchers) 6) Will it hurt if I get recommendation letters from ex-professors (who are well regarded in the field) but now work at Amazon or some other tech companies? Thinking this will need me to start prep sooner rather than later, and maybe getting back into research, otherwise start looking for good apartments in Bangalore xD
Go to Amazon India & come back in 18 months. Apply for EB1C.
It's not a bad idea - I have also been toying with the idea of potentially India FIRE (or at least FI + startup) in case things don't work out - this might be a good idea see how well I can adjust to life back home.
Coming back from India is tricky. Too much politics and bad managers. If you want to go out and come back go to Canada or EU.
That’s not enough for EB1!! Try moving to India office as a Sr. SDM (L6). Come back after a year or two to an org that can apply for EB1. Thats the only way forward
If I wait a couple of years I should be able to get 300 citations and add 2-3 patents, and maybe a few papers (based on my motivation) as well.
Unfortunately USCIS doesn’t give a f* about your citations or patents that you may get based on your “motivation”. They treat you the same way they treat an employee from Infosys. There was another post on how USCIS sent an RFE for a stanford grad. So, do what is in the rule book. Talk to an immigration lawyer(not fragoman) and help yourself.
I have done it when worked in Amazon with very very similar profile. I have a GC in hand now.
Off topic but I'd love to know why you've become disillusioned with novel research. Care to share?
TLDR: Most novel research is incremental crap that I've seen a million times before in some slightly different permutation. The few good papers that come out in a year take a significant time investment (3-6 months) which I can't do in my role while also shipping products to justify high TC. You need to be at a Deepmind/FAIR to be able to devote 6 months to a kickass research paper. If I invest in papers in the future (beyond playing with ideas with interns), it should be high impact, which isn't really possible here. Most work I do in industry is incremental which will get 15-20 citations and then be forgotten in a couple of years. Otherwise my production work has more real impact, and production is often about trading off the right level of system simplicity vs complexity which keeps future options open rather than implementing bleeding edge models that box you into a corner. Seen far too many PhDs that come in, implement super-complex novel models for simple tasks that don't need them (e.g. binary classification) to achieve marginal gains on offline metrics (that's what modern ML academia encourages the most) but practically no gains on actual downstream business metrics, but end up making production systems un-maintainable by anyone other than them. If they kept it simple, the work could have been transitioned to someone with more basic ML knowledge like an SDE and they could have moved on to a newer, hopefully more interesting projects project.
It's sad that a person with no skills from any country other than India and China has a better shot at green card than you.
Not easy. Stanford phds with 250 citations are getting rejected
Are they working for Amazon? Did you apply for EB1B?
This crazy backlog will allow universities to start selling green cards soon.
I think it is highly unlikely. I have more than 300 citations with masters degree also published in lot of top-tier venues. First and foremost you need to be invited in program committee multiple times for prestigious conferences. External reviews won’t count. You need substantial media coverage of your work and you’ll need to prove that it is of national interest. Citations are just 1 out of 8 requirements that USCIS looks at. Also I’m assuming that you are the first author on all these papers and they are published in main/research track. Unfortunately workshops won’t hold much value.
@echosmith: did you decide to apply?
I got eb1a. I have an Indian friend who has masters only and got eb1a from electrical engineering research. That was In my previous work. He doesn’t write a lot of papers. When we write papers, we often add him as a coauthor. He may publish less than 10 papers. Citations depend on areas. Note that everything you claimed needs evidence. Eg journal is top. Needs ranking. Impact needs to be national rather than regional. Eg if your work is mentioned in both nyc and ca news, you can say that is national Try to get some Recommendation letters from big name universities, such as Harvard Stanford. Purdue, ut Austin etc don’t belong to big names. Any major, assistant prof is ok if big names Try to be an associate editor in a journal. We helped our friend be AE for a Japan journal. As a Chinese, getting a PhD is only useful for green card. Nothing useful for my work
I haven't published in journals - only in top tier conferences (e.g. NeurIPS/ACL/CVPR/Interspeech) - is that going to going to be a problem - in the respective Applied ML areas, these conferences are considered pretty good
The problem with conferences is they don't have ranking and impact factors so you LL have to somehow get a citation ranking for your conference papers to prove relevance. You can also get your references to callout the conference as being one of they most important in your field and had a rejection rate of x. BTW most top tier conferences have much lower rejection rate s than top journals, so the bar is lower. I'm not sure if reviewing for conferences is considered as good as reviewing for journals but you should definitely get proof from the conference committee that you did this.