Opinion needed for Exceptional Promise profile

Stage 1
1

I would like to apply for the Exceptional Promise endorsement (I doubt that I can get the Talent route since I have 2 YOE only).
A lot of my work has been internal in the company and I’m not sure I can support my claims except with reference letters from my peers and managers.
I’ve been working as a Machine Learning Engineer at Meta in an AI team for more than 2 years now, and interned at Amazon, Adobe, and Meta. Here is my packet:

Statement of Purpose:

Still drafting, but I’ll mainly talk about wanting to found startups that use AI for social good. I’ll explain that I chose the UK because it has a mature tech hub and would like to stay in London to receive the support from Venture Capitalists and the startup ecosystem here. I also want to stay close to the various research organizations here like DeepMind, FAIR, and the universities of London.
I would also like to keep the possibility open of doing a PhD without losing the ILR track, but I’m not sure if I should mention this.

Letters of Recommendation LOR:

I have a few people in mind, but not sure which ones are eligible, and what is the top 3. Sorry the list is long so please feel free to skip to the criteria.

  1. My ex-manager and recently my manager’s manager is a leading professor in the field of Computer Vision (40k+ citations). They’ve been working with me since I joined on innovative research, but all private to the company.
  2. My new manager, less citations (>1k) but still a good h-index of 17. They’ve been my science lead since I joined.
  3. My University master thesis supervisor, an assistant professor with 12k citations. They can either recommend me or ask the head of the lab who had less interaction with me but is more renowned.
  4. My manager at Amazon in 2019, I was an intern so only 4 months of interaction. An ex-professor with ~3k citations.
  5. My manager at Adobe Research in the summer of 2017, not sure how to prove his expertise, other than their ~8 years in the company. My internship was more than 5 years ago though, does that make it irrelevant? We kept contact and he knows what I’ve been doing ever since.
  6. My previous tech lead…
  7. … and my current tech lead. I’m not sure how I can prove they are experts in the field, they’re both high-level engineers here one with 15+ years of experience in Computer Vision R&D at Automative companies and one with 6+ YOE in big tech.
  8. My internship mentor at Meta in 2018, my project is still in use at the company since 2018. Their expertise in the field is 8+ years at Meta, but I could also ask the director of that org in the company who was their manager in 2018.
  9. My second internship mentor at Meta in 2019, my project is not in use anymore, but they’re more known externally as they work in research.
    10+ I could ask a few more scientists on the team. Can one of them write me a letter or prove one of my optional criteria if they are also applying for the Global Talent visa?

Mandatory Criteria

  1. *You led the growth of a product-led digital technology company, product or team inside a digital technology company, as evidenced by reference letter(s) from leading industry expert(s) describing your work, or as evidenced by news clippings, lines of code from public repos or similar evidence.
    I joined the team as the 4th member in 2020 and we grew to ~18 people. My LOR1 can vouch for that. I helped setup the training infrastructure for our models and data pipelines.
  2. * You command a high salary or other remuneration for your services, as evidenced by commercial or employment contracts with salary information including any bonus and equity options and history of earnings.
    I believe that I do, not sure what is the minimum required. It’s one of the lowest for big tech engineers since I’m still junior. Base pay is ~70k£.
  3. * You have held or hold a significant expert role participating on panels, or individually, assessing the work of others in the same field or a field of specialisation related to the digital technology sector.
    No participation in panels, but my peers could vouch that I am the data engineering related expert in my team.

Optional Criteria

I’m not sure if I should pick two, I tried to fill all to provide backups for the weaker ones.

  1. Innovation
    a. All internal work, but we’re launching a new model to facilitate creation of some things in the Metaverse (not sure how much confidential information I can give). I came up with new ways to evaluate the model versions, and process the data for it.
    b. I also improved its overall accuracy by a few percents in successive versions. (Not sure if this goes in OC3)
    c. Is it enough though to just say that it’s in the field of Machine Learning and Augmented Reality?
    d. My work at Amazon was also internal but innovative.
  2. Work outside of my occupation
    a. Does it count that I won a hackathon as a student in 2018? https://twitter.com/START_Hack/status/967761563150057478 https://github.com/louismerlin/neo-caritas
    b. OSS I only detected a regression in Pytorch, fixed a small bug in a (800+ stars) library, and reviewed some documentation changes in Presto (a 14k stars database library)
  3. Significant Impact
    a. Built an API that scientists in our team use to experiment with their models faster. Also used by other teams to try our work.
    b. My pipelines are used to process millions of frames and the resulting datasets have led to fruitful internal research by my teammates.
    c. Point b form OC-1.
    d. My internship contributions at Meta or Adobe research are still helpful to the respective companies in some way.
    This criteria checklist states: “although you will have to demonstrate how you have made a significant impact in the sector beyond your day-to-day activities” is this a hard requirement?
  4. Academic Contributions
    a. Does internal research at the company work? My science lead can vouch that it’s on par with academic publications.
    b. My master thesis doesn’t count as per the TN website.

Additional points

Not sure if they’re relevant but could these help?

  1. My alma mater is a top ranked university (Ecole Polytechnique Federale de Lausanne)
  2. During High School I got highly placed in Math and Informatics Olympiads in Lebanon. (6th in Math)
  3. I speak 6 different languages. I only have official proof for 4.

Conclusion

Sorry for the long post, but your advice would be extremely helpful to me :frowning:
a) I would like to know who should I focus my LORs on.
b) Should I select drop some of the criteria to increase my allocated documents for others?
c) If a LOR supports one of the criteria, should I just use that?
d) I would like to avoid sharing confidential information, that my managers wouldn’t want to sign off on, is that okay? For example the list of evidences state things like “Documentation on product designs or architecture diagram clearly showing your contribution”.
e) Do I have a chance for exceptional talent? Should I wait a year, gather more contributions, and apply for it instead?

1 Like
2

Taking a look at the latest guidance from the official visa guide - Global Talent Visa | Tech Nation
I’d base my selection on 2 criteria

  1. Position in food chain aka role
  2. Expertise / citations / reputation

Also I’d filter out those with <12 months of collaboration

With that said, I’d choose #1, head of lab at university and 1 among other 7 options.

My understanding is that >5 yr ago is fine as far as “recommendation letter” is concerned. For evidences on mandatory/optional criteria, documents should be <5 yr ago.
Discussed here - Recommender from employer more than 5 years ago

I haven’t applied yet myself but from reading here - You can only choose 2 out of 4 [by design]. So yes you should drop criteria.

Yes sure. But that’s not enough for satisfying the criteria. You have to still add 2-3 documents to beef up that criteria.

Should be okay. Only write what’s possible while respecting the “confidentiality rules” you’ve signed. I remember reading someone anonymizing things or just giving high-level information in such cases. There might be smarter ways of retaining confidentiality but still proving. Unfortunately I don’t know how to do that myself.

Not an expert. But >5 YOE is the main difference between talent & promise from the official guide. So by that definition, you’d have to wait 3 years [given your current 2YOE].

1 Like
3

Thanks a lot @Chaitanya_Bapat for your input! I’ll pick
#1, my university lab research lead (and if they don’t accept, the assistant professor), and #5.

I’ll pick the Optional Criteria #1 and #3. If I have enough documents left, I’ll add the Hackathon Win as an Innovation.

I haven’t applied yet myself

Good luck to yourself if you’re going to apply

I’ll update this thread after I apply. If anyone has more tips and opinions, please feel free to share it too. Any validation is going to help me immensely.