EB1A Case Study: VP of Engineering Without Publications

This case study is anonymized. Client details, employer information, and identifying specifics have been modified to protect confidentiality. The workflow, timeline, and strategic decisions accurately reflect the preparation methodology.

Non-academic EB1A petitions require a different strategic framework than researcher cases. When a client has no publications, no academic awards, and no peer review history, the attorney must build the extraordinary ability argument from organizational evidence, market compensation, and technical contributions documented through industry rather than academic channels. This case study shows how that argument is constructed — and why one criterion required a second attempt. The EB1A regulatory standard is defined under 8 CFR 204.5(h)(3); the USCIS Policy Manual, Volume 6, Part F, Chapter 2 provides adjudicator guidance on applying the criteria to non-academic petitioners.

The Client Profile

A VP of Engineering — call him Mr. B — at a Series B AI infrastructure startup in the process of scaling toward a Series C. His professional record at the time of filing:

Organizational role: VP of Engineering for 3 years at a startup that had raised $85M Series B from recognized tier-1 venture capital firms. Led a team of 60 engineers building a distributed model training system used by multiple Fortune 500 companies. Reported to the CEO; managed 3 engineering directors.

Technical contribution: Primary author of an open-source training optimization library with 8,500 GitHub stars, 340 forks, and 1.2M downloads in the trailing 12 months. Used by multiple named AI research organizations and the basis for a funded startup's core product.

Compensation: $620,000 total (base $350,000 + target bonus $65,000 + annualized RSU value $205,000).

Conference and public presence: Talks at MLSys, NeurIPS Expo, and QCon. Regular speaker at technical conferences, primarily on distributed training and infrastructure.

What was absent: No peer-reviewed publications. No academic awards. No history of peer review or grant panel service.

The attorney's first task was identifying what was available to work with, and what was not.

Criteria Analysis

After uploading 95 documents and reviewing the AI-generated criteria mapping, the strategic picture was clear:

Strategy decision: Build on Criteria 8, 9, and 5. Accept that Criterion 5 needed more work — the initial evidence was promising but not strong enough to carry the criterion alone. Defer filing until additional expert letters and organizational adoption documentation could be added.

Building Each Criterion

Criterion 8 — Critical Role in Distinguished Organization

The organizational evidence package:

Role documentation. HR-issued organizational chart showing the VP's position: CEO at top, VP Engineering reporting directly to CEO, 3 engineering directors and 60+ engineers below. Job description documenting responsibility for the training system that constituted the company's core product.

Critical function evidence. CEO declaration letter explaining that the VP had made the core architectural decisions for the distributed training system — the product differentiating the company from competitors — and that the technical direction the VP established had been the basis for the company's most significant customer relationships.

Organizational distinction documentation. TechCrunch article about the Series B round (naming the investors and the round size), investor portfolio pages confirming the investment relationship, Fortune 500 client relationships documented anonymously ("a leading global financial services company"), and LinkedIn company profile showing 400+ employees.

Criterion 9 — High Salary

Compensation documentation. Current year's most recent pay stub (base salary), equity grant agreements (RSU grant amount, vesting schedule, grant date price), and total compensation calculation:

Comparison analysis. BLS OES data for "Computer and Information Research Scientists" showing the 90th percentile at $208,000 nationally. Levels.fyi data for "VP Engineering / Staff Engineer" at Series B technology companies showing $620,000 at approximately the 93rd percentile. The BLS comparison used the broader occupation category (conservative); the Levels.fyi comparison used company-stage-matched peers (more specific).

Criterion 5 — Original Contributions

The open-source library. The training optimization library was documented through GitHub repository analytics: 8,500 stars, 340 forks, 92 unique external contributors, 1.2M total downloads, and active community adoption with weekly issues and pull requests from unaffiliated contributors.

Named organizational adoption. Three specific named instances of organizational adoption: (1) a funded AI startup whose entire training architecture was built on the library (documented by their TechCrunch funding announcement naming the library as their core infrastructure); (2) a major AI research laboratory's open-source documentation explicitly listing the library as a dependency; (3) a university research group's published code repository using the library for their published experiments.

Expert letters. Three independent expert letters from ML practitioners: an AI research lead at a recognized technology company, an ML infrastructure architect at a large enterprise company, and a faculty member at a recognized university who had used the library in graduate course assignments. Each letter described the specific technical problem the library solved, why the approach was original, and how they or their organization had adopted it.

The RFE

The petition was filed and premium processed. USCIS issued an RFE 3 weeks after filing:

RFE ground: "The record does not establish that the petitioner's open-source library contribution constitutes a contribution of major significance to the field. The download statistics submitted represent platform-reported metrics and do not independently verify the significance or originality of the contribution."

This was a fair challenge. Download statistics are self-reported and unverified. USCIS has become increasingly skeptical of raw platform metrics as evidence of significance.

RFE response (prepared in 5 days using Immigration Copilot):

Three new independent expert letters. The attorney identified three additional independent experts through the client's professional network: a founding engineer at a company that had integrated the library into their production infrastructure, an ML platform engineer at a large technology company who had evaluated and adopted the library, and a PhD researcher whose published work cited the library. Each letter was briefed using the same methodology as the original letters, with emphasis on describing specific use cases and the specific problem the library solved — not the download statistics.

Named organizational evidence. Documentation that a named funded startup had built its core product on the library: the startup's public documentation listing the library as a core dependency, combined with their Series A announcement press release. This provided a named, verifiable example of a third party building commercial value on the petitioner's contribution.

Technical community adoption evidence. A published blog post by a Google AI researcher describing their team's implementation of the approach, naming the library as the inspiration. A conference workshop paper citing the library in its methodology section.

Result: Approved on RFE response. No second RFE.

Key Lessons

Non-traditional EB1A requires more expert letter investment up front. The RFE came because the initial Criterion 5 argument rested on adoption metrics without enough independent expert voices documenting the technical significance. Six expert letters instead of three, sourced from users with named organizational affiliations, would likely have prevented the RFE.

Step 2 framing for business-track cases requires explicit attorney drafting. The petition brief must articulate why the combined evidence — critical role at a distinguished company, top-percentile compensation, and adopted technical contribution — establishes that this person is among the small percentage at the very top of their field in AI infrastructure specifically. That argument requires a coherent narrative that connects the three criteria into a unified extraordinary ability claim.

AI assistance is as valuable for RFE response as for initial filing. Having the structured KB available meant the attorney could work from the organized case record — not re-read 95 documents — when preparing the RFE response. The new expert letters were drafted from the same KB facts that had generated the original letters, updated with additional adoption evidence gathered after the RFE.

For the complete criteria strategy for technology professionals and executives, see EB1A evidence strategy by client profile. For how to prevent the Criterion 5 RFE pattern that this case encountered, see the EB1A RFE prevention playbook. For the business-track criteria framework and how to combine Criteria 5, 8, and 9 for non-academic professionals, see EB1A without publications: what evidence works.

Immigration Copilot handles document classification, criteria mapping, and petition generation for technology executive EB1A cases. Get started →