A real sample dossier
Generated for Dr. Priya Raman, a fictional ML-researcher-turned-founder persona (Dr. Priya Raman is not a real person), generated in the exact shape the production pipeline emits — the same intake schema, the same validator gate, the same attestation stamp. It is an example artifact, not a statement about a real applicant.
This is exactly what a paying customer receives — the same pipeline shape, the same validator gate, the same rendering — shown in full. Every document carries its attestation line; criteria with no evidence are rated Absent, not inflated. The letter briefs are skeletons with bracketed placeholders, never ghost-written letters.
Cover Note to the Reviewing Attorney
Based on evidence provided by Dr. Priya Raman on 2026-07-29. Prepared by an automated evidence-research service, not a law firm; not legal advice; no visa-route recommendation; no USCIS outcome prediction; every claim requires independent verification by the reviewing attorney.
Every factual claim cites the applicant's intake and carries verification hooks naming what you would need to independently verify it. The strength ratings are evidence-availability ratings only: they describe how much evidence the intake supplied for each criterion, never eligibility. Two criteria are rated Absent and one Thin; the gap analysis names the kind of evidence that would close each.
The recommendation-letter briefs are skeletons with bracketed placeholders — the recommenders must add their own personal knowledge and opinions; nothing in them is a drafted opinion. The applicant's delivery contact is priya.raman@example.com.
Applicant Profile Summary
Based on evidence provided by Dr. Priya Raman on 2026-07-29. Prepared by an automated evidence-research service, not a law firm; not legal advice; no visa-route recommendation; no USCIS outcome prediction; every claim requires independent verification by the reviewing attorney.
Dr. Priya Raman is a machine-learning researcher turned founder working in machine-learning infrastructure. She holds a PhD in Computer Science from ETH Zürich (2019), where her doctoral work focused on distributed training of large neural networks.
From 2019 to 2023 she was a Research Scientist at Google DeepMind (Zürich), where she co-authored the FLAN paper (Scaling Instruction-Finetuned Language Models), cited 4,000+ times. Her publication record covers 8 peer-reviewed papers at NeurIPS, ICML, and ACL with ~5,800 total citations and an h-index of 16.
Since 2023 she has been Co-founder & CTO of an 8-person ML infrastructure startup (YC W24) that raised a $4.5M seed round. The company builds distributed-training infrastructure that lets small research teams train large models without hyperscaler budgets, serving university labs and early-stage startups. She is a core contributor to a distributed-training library with 12k GitHub stars and maintains two personal open-source projects with ~1.5k stars combined. She reviews for NeurIPS and ICML at an area-chair-adjacent load of 15+ papers per year.
Criteria Mapping — Dr. Priya Raman (O-1A extraordinary ability)
Based on evidence provided by Dr. Priya Raman on 2026-07-29. Prepared by an automated evidence-research service, not a law firm; not legal advice; no visa-route recommendation; no USCIS outcome prediction; every claim requires independent verification by the reviewing attorney.
Strength ratings describe evidence availability only — how much the intake supplied — not eligibility and not any outcome prediction.
Awards or prizes for excellence (8 CFR 214.2(o)(3)(iii)(A))
Evidence from intake: Best-paper honorable mention, ACL 2022 workshop. Strength: Thin Verification hooks: Official conference award notification or certificate; conference records showing the award and its selection process.
Membership in associations requiring outstanding achievement (8 CFR 214.2(o)(3)(iii)(B))
Evidence from intake: None provided in intake. Strength: Absent Verification hooks: N/A — no evidence supplied.
Published material about the applicant (8 CFR 214.2(o)(3)(iii)(C))
Evidence from intake: Quoted in TechCrunch coverage of the company's seed round (2024); invited guest on a well-known ML podcast (2024), ~50k listeners. Strength: Moderate Verification hooks: TechCrunch article URL and archived copy; podcast episode link and audience metrics.
Judging the work of others (8 CFR 214.2(o)(3)(iii)(D))
Evidence from intake: Reviewer for NeurIPS 2022–2024 and ICML 2023–2024; area-chair-adjacent load of 15+ papers per year. Strength: Strong Verification hooks: Reviewer invitation emails, assignment records, and review-load documentation from the conferences.
Original contributions of major significance (8 CFR 214.2(o)(3)(iii)(E))
Evidence from intake: Co-author of the FLAN paper (Scaling Instruction-Finetuned Language Models), cited 4,000+ times; core contributor to a distributed-training library with 12k GitHub stars used in production by research labs. Strength: Strong Verification hooks: Google Scholar citation report with timestamp; GitHub repository statistics and contribution history.
Authorship of scholarly articles (8 CFR 214.2(o)(3)(iii)(F))
Evidence from intake: 8 peer-reviewed papers at NeurIPS, ICML, and ACL (2019–2024); ~5,800 total citations, h-index 16. Strength: Strong Verification hooks: Complete publication list with venues; Google Scholar profile export.
Critical role for distinguished organizations (8 CFR 214.2(o)(3)(iii)(G))
Evidence from intake: Co-founder & CTO of an 8-person ML infrastructure startup (YC W24, $4.5M seed) — owns the core training-infrastructure architecture the product is built on; Research Scientist at Google DeepMind (2019–2023) on the FLAN project. Strength: Moderate Verification hooks: Employment verification letters; incorporation and funding documentation; role descriptions.
High salary or remuneration (8 CFR 214.2(o)(3)(iii)(H))
Evidence from intake: None provided in intake. Strength: Absent Verification hooks: N/A — no evidence supplied.
Honest Gap Analysis
Based on evidence provided by Dr. Priya Raman on 2026-07-29. Prepared by an automated evidence-research service, not a law firm; not legal advice; no visa-route recommendation; no USCIS outcome prediction; every claim requires independent verification by the reviewing attorney.
This analysis names where the intake record is thin or absent and the KIND of evidence that would close each gap. It describes evidence types only — not chances of success, and not which route to file.
Thin criteria
Awards or prizes for excellence. The only award in the record is a workshop best-paper honorable mention (ACL 2022 workshop). The kind of evidence that would close this gap: major-conference awards, fellowship-grade honors, or industry-recognized prizes with documented selectivity.
Absent criteria
Membership in associations requiring outstanding achievement. No memberships were provided in the intake. The kind of evidence that would close this gap: membership in an association whose admission requires outstanding achievement judged by recognized experts, with the admission criteria documented.
High salary or remuneration. No compensation information was provided in the intake (the applicant chose not to include it). The kind of evidence that would close this gap: compensation documentation paired with field- and geography-appropriate salary survey data.
Intake items marked "not yet"
- Patents: none provided — patent filings, if any exist, would add to the original-contributions row.
- Memberships: none provided.
- Salary / compensation: not supplied by the applicant.
The core of the record — judging, original contributions, and scholarly articles — is strong. Two criteria are absent and one is thin; the dossier maps what exists rather than inflating what does not.
Evidence Register — Dr. Priya Raman
Based on evidence provided by Dr. Priya Raman on 2026-07-29. Prepared by an automated evidence-research service, not a law firm; not legal advice; no visa-route recommendation; no USCIS outcome prediction; every claim requires independent verification by the reviewing attorney.
1. PhD in Computer Science, ETH Zürich (2019) — supports Authorship of scholarly articles — verify via: diploma and thesis documentation.
2. FLAN paper citation count (4,000+) — supports Original contributions of major significance — verify via: Google Scholar citation report with timestamp.
3. Best-paper honorable mention, ACL 2022 workshop — supports Awards or prizes for excellence — verify via: conference award notification or certificate.
4. 8 peer-reviewed papers at NeurIPS/ICML/ACL (incl. arXiv:2210.11416), ~5,800 citations, h-index 16 — supports Authorship of scholarly articles — verify via: Google Scholar profile export.
5. NeurIPS 2022–2024 and ICML 2023–2024 reviewer service, 15+ papers/year — supports Judging the work of others — verify via: reviewer invitation emails and assignment records.
6. TechCrunch quote in seed-round coverage (2024) — supports Published material about the applicant — verify via: article URL and archived copy.
7. ML podcast appearance, ~50k listeners (2024) — supports Published material about the applicant — verify via: episode link and audience metrics.
8. Research Scientist, Google DeepMind (2019–2023) — supports Critical role for distinguished organizations — verify via: employment verification letter.
9. Co-founder & CTO, ML infrastructure startup (YC W24, $4.5M seed) — supports Critical role for distinguished organizations — verify via: incorporation documents and funding announcement.
10. Core contributor, distributed-training library (12k GitHub stars) — supports Original contributions of major significance — verify via: GitHub repository statistics and contribution history.
Recommendation-Letter Briefs — Dr. Priya Raman
Based on evidence provided by Dr. Priya Raman on 2026-07-29. Prepared by an automated evidence-research service, not a law firm; not legal advice; no visa-route recommendation; no USCIS outcome prediction; every claim requires independent verification by the reviewing attorney.
Each brief is a skeleton: factual anchors from the intake the recommender could plausibly speak to, plus bracketed placeholders where only the recommender can add their own assessment. These are not letters; the recommender writes the letter.
Brief 1 — Prof. Andreas Keller
Relationship to applicant: PhD thesis advisor, ETH Zürich. Factual anchors the letter can cover:
- Doctoral work on distributed training of large neural networks (doctoral researcher 2015–2019).
- Subsequent publication record: 8 peer-reviewed papers at NeurIPS, ICML, and ACL, ~5,800 total citations, h-index 16. [For the recommender to complete]: [your personal assessment of the thesis quality and originality] [specific examples of the applicant's research approach you observed directly] [how the applicant compared to other doctoral students you have supervised] [your knowledge of the field-level impact of the work]
Brief 2 — Dr. Sofia Marchetti
Relationship to applicant: former research lead at Google DeepMind, FLAN co-author. Factual anchors the letter can cover:
- The applicant's role as Research Scientist at Google DeepMind (Zürich), 2019–2023.
- Co-authorship of the FLAN paper (Scaling Instruction-Finetuned Language Models), cited 4,000+ times. [For the recommender to complete]: [the applicant's specific contributions to the FLAN project that you witnessed] [your assessment of the applicant's technical abilities relative to peers at the organization] [examples of research leadership or initiative you observed]
Brief 3 — James Wu
Relationship to applicant: CEO of a partner company using the startup's training infrastructure in production. Factual anchors the letter can cover:
- The applicant's current role as Co-founder & CTO of the ML infrastructure startup (YC W24, $4.5M seed).
- The partner company's production use of the training infrastructure the applicant architects. [For the recommender to complete]: [the concrete value the infrastructure provides to your company] [your assessment of the technology's significance for teams like yours] [why you chose this product over alternatives you evaluated]