EB-2 National Interest Waiver · Self-Petition

Your NIW petition,
start to finish.
Evaluation to filing, in one workspace.

Evaluate your case. Organize the evidence record. Draft petition materials. Prepare for RFE response. Papers supports the full NIW lifecycle — from first assessment to final filing.

Free · No credit card · ~5 min

EB-2 NIW Evaluation·Three-Prong Analysis
Analyzing…
Potentially a promising case

Prong 1

Merit & Importance

Very Strong

Prong 2

Well-Positioned

Needs Development

Prong 3

Benefit of Waiver

Promising

Summary: Filing today would likely draw an RFE on Prong 2. With ~6 months of preparation, this case can become potentially strong.

01

Decide

See exactly where your case stands — prong by prong — before you invest time preparing.

02

Prepare

Organize every exhibit. Papers finds official policy sources that anchor your work's national importance.

03

Draft

Petition letter and recommendation letters written from your actual evidence record, not placeholder language.

04

Respond

If USCIS issues an RFE, Papers maps each issue to the criterion being questioned and drafts the response.

01 · Decide

First, see whether NIW is worth pursuing.

Papers turns your background into an NIW case map: the work you want USCIS to evaluate, why it matters nationally, what evidence supports it, and what is still missing.

Define the work clearly

NIW is not just about your resume. It is about the specific work you plan to advance in the United States.

Test the three requirements

Papers checks national importance, whether you are positioned to do the work, and why a job offer should be waived.

Surface likely USCIS questions

The report shows where an adjudicator may ask for more evidence before you spend time drafting.

Prong 2·Well-Positioned to Advance
Needs Development

Proposed Endeavor

"Developing AncestralBERT — a genomic AI framework for rare disease diagnosis across under-represented populations."

Policy Manual sub-factors

Education, skills & knowledge

PhD Computational Genomics, Stanford. 8 years specialty experience.

Very StrongDocumented
Record of success

47 independent citations across 2 papers. No external adoption letters.

Needs DevelopmentUnsupported
Model or plan

Endeavor narrative articulated; scaling mechanism to third parties needs detail.

PromisingInferred
Progress to date

Initial model published and open-sourced. No U.S. clinical deployment yet.

PromisingInferred
Stakeholder interest

No independent interest from U.S. clinical adopters or grant agencies on record.

High RiskUnsupported

02 · Prepare

Organize your evidence. Find what you are missing.

The evidence board tracks everything your petition needs — publications, awards, credentials, and national impact evidence. Connect Google Scholar to sync your publications automatically. Papers also helps find official policy sources that anchor your work's national importance.

Google Scholar import

Publications, citation count, and h-index synced automatically. No manual entry needed.

National evidence finder

Live search across OSTP, DOE, NSF, and DHS — finds official policy documents that support your specific work.

AI guidance on what matters

Papers tells you which evidence categories are weak for your profile — academic, industry, or entrepreneur — and what to add.

Case Builder/Evidence · 46 exhibits
Letters · 4
AI legal-reasoning research · National Importance

Three federal sources strengthen your Prong 1 argument:

NIH Strategic Plan for Data Science (2023) — names rare-disease diagnostics as an AI research priority.

NSTC Critical & Emerging Technologies List — AI and Biotechnology designated.

GAO-23-105714 — quantifies the diagnostic burden across 30M Americans.

Field-Influencing Research7 exhibits
Sharma et al. — “AncestralBERT: variant classification across diverse genomes” (Nature Genetics, 2025)23 citations
Sharma et al. — “Closing the diagnostic gap for under-represented ancestries” (Cell, 2024)24 citations

03 · Draft

Petition letter and recommendation letters, drafted end to end.

Papers drafts each section of the petition letter around your actual evidence — not placeholder language. Recommendation letters are tailored to each recommender's relationship to the work, with guidance on independence and evidentiary value.

B
I
U
H1
H2
H3
Generate

Section 3. Dr. Sharma is well-positioned to advance the proposed endeavor.

3.1 Training and expertise.

Dr. Sharma earned her Ph.D. in Computational Genomics from Stanford University in 2019. Over eight years she developed the AncestralBERT framework — a transformer-based approach to variant classification across ancestrally diverse genomes — placing her among a small group of researchers with expertise spanning genomic sequencing, deep learning, and rare disease phenotyping.

3.2 Record of success.

Dr. Sharma has published 14 peer-reviewed articles, with 47 independent citations across two first-author publications in Nature Genetics and Cell. Leading researchers have independently assessed her contributions:

"Dr. Sharma's contributions to ancestral genome modeling are substantial. Her AncestralBERT framework has been adopted in our laboratory research protocols, and her technical depth across genomics and machine learning is exceptional among peers at her career stage."

(Exhibit 14, Letter from Prof. J. Mitchell, MIT CSAIL)

"The AncestralBERT framework reduces variant classification error by 23% across under-represented populations — addressing one of the most persistent challenges in rare disease diagnostics. Dr. Sharma is uniquely positioned to extend this work in the United States."

(Exhibit 15, Letter from Dr. S. Liu, Harvard Medical School)

3.3 Recognition by the field.

Dr. Sharma was invited to deliver a keynote at NeurIPS 2024 on diagnostic equity in genomic AI (Exhibit 19) and was named an ISCB Fellow in recognition of her contributions to computational biology (Exhibit 20). These distinctions reflect independent expert judgment that her work has advanced the field.

Petition built from your evidence

Every section cites your uploaded exhibits. AI flags which evidence supports which claim in the petition.

Recommender independence matters

Letters from close collaborators carry less weight. Papers tracks the mix and advises who to ask for independent letters.

Profile-specific drafting

Academic, industry, and entrepreneur profiles get different argument structure based on what the evidence record looks like.

04 · Respond

An RFE is a specific challenge. Papers gives a specific answer.

Papers reads the notice, identifies which criterion is being questioned, and turns each issue into a response strategy with the evidence you need to add.

Maps every issue to a criterion

Each paragraph of the RFE is linked to the specific criterion being questioned — so nothing falls through the gaps.

Specific evidence to add, not generic advice

Papers tells you what type of evidence addresses each objection — expert letters, policy citations, adoption data — and why.

Time is limited — act on the right things

USCIS gives a fixed response window. Papers prioritizes the issues by severity so you spend time on what matters most.

RFE Analysis·3 issues identified
High RiskProng 3 · Benefit of Waiver

Failure to show waiver benefit outweighs labor certification.

No evidence U.S. genomic medicine centers cannot find qualified domestic candidates. Adjudicators commonly cite this gap in Prong 3 denials.

Needs DevelopmentProng 1 · Merit & National Importance

Endeavor — not the field — must show national importance.

Currently framed at field level (AI / genomics). USCIS adjudicates the specific endeavor. No evidence the framework has been adopted by U.S. clinical institutions.

Needs DevelopmentProng 2 · Well-Positioned

Insufficient record of success in the field.

47 citations across 2 papers is borderline. Additional field-influence evidence is typically requested.

Why Papers is different

Other tools teach AI what good petitions look like. We built ours to reason from the standard they have to meet.

Any tool can prompt AI with sample petitions and call it legal intelligence. A practicing attorney reads the output and immediately sees what's off. A self-petitioner cannot — they can only trust what the AI produced.

"When a professional uses AI and it makes a subtle error, they catch it. When a self-petitioner does, they file it."

Standard approach

Prompt AI with sample petitions and style guidelines

Generate output calibrated to what approved petitions look like

Errors visible to professionals. Not to self-petitioners.

Output: looks right. May not be right.

Papers approach

01

Define the specific endeavor USCIS will evaluate

02

Test each criterion from the adjudicator's perspective

03

Tag every claim: Documented, Inferred, or Unsupported

04

Draft only what the record actually supports

Output: right because the reasoning is right.

Assist by Attorney

AI prepares the case file. Attorneys apply judgment when you want it.

The serious path is not AI versus attorneys. Papers lets you prepare the case with structure, then bring a licensed immigration attorney into the exact part of the lifecycle where professional judgment matters.

Assist by Attorney·How it works
Coming soon
01

Evaluate & prepare

You + Papers AI

Case assessment, evidence organization, petition draft built from your record.

02

Request counsel

Your choice, any stage

Share the full case file with a licensed immigration attorney at whatever point you want judgment.

03

Counsel reviews

Licensed attorney

Applies judgment on evidence gaps, prong theory, and filing strategy. AI prepares; attorneys decide.

The complete case file — evaluation, evidence map, petition draft — transfers to counsel at whatever stage you choose.

Claude plugin and skills

Inspect the reasoning, not just the output.

The NIW evaluation method is available as an open-source Claude plugin and skill. Most applicants will use the hosted product, but making the reasoning inspectable shows how seriously we treat responsible legal AI.

Published plugin, public method

The Claude plugin exposes the reasoning structure behind the evaluation: work definition, evidence basis, NIW criteria, and likely USCIS questions.

Product is still more than a prompt

The hosted workspace adds documents, case state, drafting, evidence organization, RFE planning, and attorney handoff.

View the open-source skill on GitHub

thepapers-niw · niw-evaluate

$ claude --plugin-dir ./niw-skills

/thepapers-niw:niw-evaluate

Welcome. I evaluate EB-2 NIW cases

  against the Dhanasar standard — prong

  by prong, with an honest read on

  where your record stands today.

Share any of the following:

  – A CV or LinkedIn-style summary

  – Your field, country of birth, timing

  – A proposed endeavor (if any)

What would you like to share?

thepaperscompany/niw-skillsPublic · Apache 2.0

FAQ

Common questions.

More questions? Contact us.

EB-2 National Interest Waiver

Prepare your NIW petition from eligibility to filing.

Start with a free evaluation. Then continue into the case workspace to organize evidence, draft petition materials, prepare for RFE challenges, and bring in attorney review when needed.

Free · No credit card required

Immigration attorney? Explore partnershipHave questions? Contact us