The Epidemiologist Who Saw the Outbreak Coming, and Almost Lost His Place in America

Dr. Rohan Chakraborty, a healthcare data scientist trained as an epidemiologist, built real time disease surveillance systems used by state health departments across the country. His National Interest Waiver petition triggered a Request for Evidence that questioned all three Dhanasar prongs at once. It was approved.

1. A Notice in the Middle of Flu Season

When the Letter Came While the Wards Were Filling

It was the second week of January when the Request for Evidence arrived in his inbox, and Dr. Rohan Chakraborty almost laughed at the timing.

Earlier that morning, before he had even finished his coffee, his dashboard had already flagged anomalous emergency department visit patterns in three counties across two states. Influenza A activity was running about ten days ahead of the seasonal forecast. RSV was spiking in pediatric units in the Pacific Northwest. A small but unusual cluster of severe respiratory illness in older adults had appeared in a rural Midwestern county, and the model he had built was flagging it for human review.

He had spent the last four hours on a video call with a state epidemiologist trying to figure out whether the cluster was a statistical artefact or something that needed investigation. They had decided it needed investigation.

Then he opened his email.

The subject line read: Request for Evidence, I-140, EB-2 National Interest Waiver.

He read the document twice. Eighteen pages, three separate challenges, each one complicated on its own. Together, they felt like a wall.

He had spent the morning teaching a machine to spot outbreaks before they happened. Now an outbreak was happening to him.

The U.S. Citizenship and Immigration Services adjudicator, in measured bureaucratic prose, was asking him to prove that his work mattered, that he was the right person to do it, and that the country would be better off if it skipped its usual hiring rules and just let him stay.

He had spent six years building tools that helped public health agencies see disease outbreaks coming days, sometimes weeks, before they were visible to the human eye. He had a PhD in Epidemiology from a top twenty American program. He had transitioned into industry to scale his work, and his models were now embedded in the surveillance infrastructure of more than a dozen state health departments.

And the federal government, the same federal government whose Centers for Disease Control was using his data, wanted him to explain himself.

He had built the warning system. The warning had come for him.

When his file landed at Visa Architect a few days later, I read it twice and then said something my associates would later quote back to me on harder cases: ‘The work is here. We just have to make it findable.’

2. The Quiet Revolution in Public Health Data

How an Old Field Met New Math

To understand the complexity of Dr. Chakraborty's case, you have to understand what has happened to public health surveillance over the last decade.

For most of the twentieth century, infectious disease surveillance in America worked the same way it had since John Snow mapped cholera deaths in Victorian London. Doctors saw patients. Patients had symptoms. Reports were filed. Cases were counted. The data made its way, often slowly, to county and state health departments. Trends became visible only after enough cases had accumulated to form a pattern. By the time public health officials could see an outbreak, it had usually already happened.

Then computing changed all of that.

Electronic health records became standard. Pharmacy data went digital. Emergency departments started capturing chief complaints in structured fields. School absence data became aggregable in real time. Wastewater testing emerged as a leading indicator. Suddenly, the limiting factor in disease surveillance was not data availability. It was data interpretation.

For two centuries, public health had counted the dead. Now it was trying to count the almost dying. That shift, more than any single vaccine or drug, was what people like Dr. Chakraborty had been quietly building.

This is where a new kind of professional emerged, someone who was not quite an epidemiologist in the traditional sense and not quite a data scientist in the Silicon Valley sense. Someone who understood both the messy biology of disease transmission and the mathematics of statistical learning. Someone who could look at a stream of millions of structured and unstructured data points and find the signal of an emerging outbreak inside the noise of ordinary human illness.

Dr. Chakraborty was one of these people.

He is a composite character, drawn from patterns that immigration attorneys at Visa Architect have seen repeatedly across public health and data science professionals. The technical and procedural details of his case, however, reflect the real world in unsettling detail.

He had earned his PhD in Epidemiology with a methodological focus on Bayesian hierarchical models for disease forecasting. His dissertation had been on early warning systems for respiratory viral outbreaks. After a brief postdoctoral fellowship, he had taken what some of his academic mentors considered an unusual step. He left the university for a healthcare analytics company that worked directly with state and local health departments, integrating clinical data streams to build real time surveillance dashboards.

He had not left academia because he stopped caring about science. He had left because, as he sometimes put it in conversation, "I wanted to build the thing, not just write papers about how someone should build the thing."

3. The Work That Was Hard to See

A Scientist Whose Best Work Lived Inside Hospital Firewalls

When Dr. Chakraborty came to Visa Architect, the first conversation took close to three hours. Not because his story was confusing, but because the parts of it that mattered most were also the parts that were hardest to document.

His resume looked, on the surface, modest.

He had ten peer reviewed publications. That was strong for someone seven years out of his PhD, but not the seventy or hundred that a tenured academic of similar age might have. His total citation count was around 240, which was respectable for an industry data scientist and unremarkable for an academic epidemiologist. He had an h index of 8.

He had no major awards. He had received a Rising Investigator recognition from a regional epidemiology society as a postdoc, and his employer had given him an internal innovation prize, but he had nothing of national stature. He had never been on television. He had been quoted in two trade publications and one CDC adjacent newsletter.

By the metrics that immigration adjudicators are trained to recognize, he looked like a competent mid-career professional. By the metrics that mattered in his actual field, he was something else entirely.

His core work over the last five years has been a real-time syndromic surveillance platform. The system ingested data from emergency department visits, urgent care chief complaints, outpatient diagnostic codes, lab orders, retail pharmacy purchases, and school absence reports across more than 1,800 hospitals and clinics. It used a Bayesian ensemble model he had designed to detect statistical anomalies in disease activity, ranging from ordinary seasonal influenza to potential novel pathogen emergence.

The platform was now in production use at fourteen state health departments and three large city public health agencies. It had been used to detect early signals of multiple recent disease events, including an unusual mpox cluster in 2022, an early RSV surge in pediatric populations in 2023, and several localized norovirus outbreaks. State epidemiologists had cited it in formal incident response documentation. CDC field officers had relied on its outputs during multiple emergency activations.

He had also built, on a separate research track, a health disparities module that overlaid disease surveillance data with social determinants of health indicators. The module flagged patterns where particular communities were experiencing disproportionate disease burden, often before traditional surveillance methods would have detected the inequity. State health departments in two of the largest U.S. states had used the module to redirect resources and adjust public health messaging.

The problem, from an immigration standpoint, was visibility.

His syndromic surveillance work was deployed inside government and hospital firewalls. The actual algorithms were trade secrets. The performance metrics were owned by his employer. The outputs were used in incident response, not published in journals. The state epidemiologists who relied on his system trusted him personally, but their official documentation rarely named individual contributors at vendor companies.

He had built the kind of infrastructure that, when it works, becomes invisible. People do not notice the outbreaks that did not happen. They notice the ones that did. And the work that prevents the second category from becoming visible tends to disappear into the routine functioning of public health.

Nobody writes a thank you note for the disaster that did not happen.

For an immigration case, that invisibility was the central problem.

4. Why the National Interest Waiver Was the Right Path

Choosing the Right Door, Not the Most Glamorous One

There is, in this work, a temptation that Visa Architect resists deliberately, which is to push every accomplished applicant toward the EB-1A "extraordinary ability" category. The EB-1A is faster. It bypasses the green card backlog that affects India born applicants most painfully. It carries a certain prestige that resonates with high achieving professionals who are used to being told they are exceptional.

For Dr. Chakraborty, after careful evaluation, the recommendation was different. The right pathway was the National Interest Waiver under the EB-2 category.

The reasoning had nothing to do with modesty. It was about strategy.

The EB-1A requires meeting at least three of ten specific criteria, and each one needs to be defensible against an adjudicator who will scrutinize claims aggressively. Dr. Chakraborty's profile, while substantively impressive, was thin on the kinds of evidence that EB-1A criteria typically demand. His publication record was modest. His awards were regional. His media presence was limited. The original contributions criterion was strong but would have to carry an outsized share of the case.

The National Interest Waiver took a different angle. It did not ask whether he was at the top of his field globally. It asked three different questions, the framework set out in the 2016 administrative decision Matter of Dhanasar.

First, does the proposed endeavor have substantial merit and national importance?

Second, is the petitioner well-positioned to advance the proposed endeavor?

Third, would it, on balance, benefit the United States to waive the labor certification requirement?

For someone whose work directly served U.S. public health agencies, who had documentary evidence of state and federal use, and whose field was self evidently in the national interest, the NIW framework fit the shape of the case naturally.

The EB-1A would have asked Dr. Chakraborty to prove he was a star. The NIW would let him prove what was actually true and arguably more important: that his work mattered to the country, that he was the right person to do it, and that forcing him through standard labor certification would damage something the country needed.

As I put it later, when explaining the decision to a group of new associates: ‘EB-1A is a vanity test. NIW is a value test. Most of the engineers and scientists we see are stronger on the second one. Our job is to know the difference.’

The decision was made. The petition would be filed under the National Interest Waiver. That decision turned out to be the right one. It also turned out, by no means, to be the easy one.

5. Building a Case Around Invisible Work

Turning Trade Secrets and Government Contracts Into Evidence

The next four months were spent on what Visa Architect's team internally calls evidence translation, the art of taking work that is inherently difficult to document and constructing a record that an adjudicator can engage with.

The first prong, substantial merit and national importance, was the easiest to argue conceptually but required careful construction.

Public health surveillance is, almost by definition, a matter of national importance. The 2016 Dhanasar decision specifically lists improving public health among examples of national importance. Federal investments in pandemic preparedness, disease surveillance modernization, and health equity infrastructure run into the tens of billions of dollars annually. The CDC's Center for Forecasting and Outbreak Analytics, established in 2021, had been described by the agency's own leadership as a national priority.

Dr. Chakraborty's work fell directly into this priority area.

The petition opened with a foundation that an adjudicator could not reasonably dismiss. It connected his specific endeavor, building data-driven early warning systems for infectious disease outbreaks, to named federal initiatives. The CDC's Data Modernization Initiative. The HHS Pandemic and All Hazards Preparedness Act priorities. The White House's Cancer Moonshot, which had a public health surveillance component. The Healthy People 2030 objectives related to infectious disease detection and health equity.

For each connection, the petition included specific citations to federal documents, congressional reports, and agency publications. The argument was not that public health is generally important, an argument adjudicators see thousands of times a year and tend to treat as boilerplate. The argument was that this specific kind of work, real time multi source disease surveillance with health equity overlays, was a federally identified priority, and Dr. Chakraborty's endeavor was directly aligned with it.

The second prong, well positioned to advance the endeavor, was harder.

This is the prong where most NIW cases live or die. It is also the prong where industry data scientists struggle most. The question is not whether you have done good work in the past. It is whether the evidence supports a finding that you can plausibly continue to advance the endeavor in the future. That finding rests on a record of past achievement, current expertise, and future trajectory.

For Dr. Chakraborty, the team built the well positioned argument on five legs.

The first was his formal training: a PhD in Epidemiology from a top tier program, methodological specialization in disease forecasting, and postdoctoral training in real time surveillance systems. This was the foundation that any NIW case requires under the EB-2 advanced degree provision, but the team made sure to detail not just the credentials but the substantive depth of his methodological training.

The second leg was his deployed work. The syndromic surveillance platform was running at fourteen state health departments. Detailed letters from state epidemiologists at multiple agencies described, with specificity that took weeks to assemble, exactly how his system had affected their public health response. Three letters described specific outbreaks where his system had provided early warning that informed agency response. One letter, from a senior epidemiologist at a state department, described a specific norovirus outbreak in 2023 where the system's early flag had allowed the agency to issue advisories two days earlier than would otherwise have been possible.

That epidemiologist's letter included a sentence that the legal team would quote back, in slightly varied form, several times across the petition. "We did not have to wait for people to start dying to know something was wrong. The system told us first."

The third leg was his methodological output. While many of his core algorithms were proprietary, he had published methodological papers describing the general statistical frameworks he used. These had been cited in CDC technical reports and adopted by at least two academic research groups working on related problems.

The fourth leg was his health equity work. The disparities module he had built had been used in actual policy decisions. The team obtained letters from public health officials describing how the module's outputs had informed resource allocation in underserved communities. This was not abstract impact. It was specific, documented, and tied to outcomes.

The fifth leg was his trajectory: pending patent applications, new research collaborations with academic institutions, and a grant collaboration with a federal research entity. The argument was not just that he had done important work but that he was actively positioned to do more of it.

The third prong, the balancing factor, required a different kind of argument.

The labor certification process, which the NIW asks to waive, exists to protect U.S. workers. The standard test under PERM is whether a U.S. worker is available who can perform the job. To convince an adjudicator to waive this test, the petition has to establish that the petitioner's specific contributions are something the United States should not delay or risk losing.

Two arguments carried this prong.

The first was time sensitivity. Public health threats do not wait for bureaucratic timelines. The PERM process typically takes 12 to 24 months. During that period, surveillance systems would either lose continuity of expertise or have to be staffed with less qualified personnel. The risk was concrete and could be quantified by reference to specific past events where Dr. Chakraborty's work had affected outbreak response.

The second was uniqueness of role. The argument here was careful. Visa Architect does not argue, and adjudicators do not credit, claims that an applicant is irreplaceable in any absolute sense. The argument was instead that the specific intersection of skills required for this work, training in epidemiological methods, fluency in modern statistical learning, deep familiarity with U.S. public health data infrastructure, and an established working relationship with state health agencies, was rare. The petition documented this rarity through Bureau of Labor Statistics data on occupational categories, peer reviewed studies on the public health data science workforce shortage, and direct testimony from public health agency officials describing recruitment difficulties for personnel with this profile.

The petition was filed in October. It ran to roughly 1,650 pages, including all exhibits, expert letters, and supporting documentation. The legal brief alone was 73 pages.

Eight weeks later, the RFE arrived.

6. The RFE That Came in Three Parts

A Triple Challenge to the Foundation of the Case

In RFE language, what arrived was not really an inquiry. It was a stress test.

The Request for Evidence was, in the assessment of Visa Architect's senior team, one of the more substantial NIW RFEs they had encountered that quarter. Not because any single argument in it was novel, but because all three Dhanasar prongs were challenged simultaneously, and the challenges were specific enough that responses would require new evidence rather than rearguing material already submitted.

The first challenge went after national importance.

The adjudicator's framing was sophisticated. The officer acknowledged that public health surveillance, in the abstract, was a matter of national importance. But the adjudicator argued that Dr. Chakraborty's specific endeavor, building proprietary surveillance products for a private healthcare analytics company whose customers were individual state and local health agencies, was not properly characterized as national in scope. The argument was, in essence, that while the field was national, the petitioner's actual day-to-day work consisted of a series of regional or local engagements with individual agencies. The RFE asked for evidence that the work had an impact extending beyond the specific jurisdictions of the agencies that were the company's customers.

The second challenge went after the well-positioned prong, and it was the most technically demanding.

The adjudicator pointed out that much of the evidence submitted relied on letters from state epidemiologists describing specific instances where Dr. Chakraborty's system had detected an outbreak signal. The officer questioned whether the letters established that the signal detection was attributable to Dr. Chakraborty's individual contributions or to the broader product team at his company. In a corporate setting, the officer noted, software products are typically built by teams. The petitioner's role within the team needed to be more clearly differentiated. The RFE asked for evidence demonstrating which specific algorithmic, methodological, or system design contributions originated with Dr. Chakraborty and were not the work of the broader engineering team.

The third challenge was after the balancing factor.

This was, in some ways, the most difficult. The adjudicator argued that Dr. Chakraborty was a U.S. trained PhD with substantial U.S. work experience. The implication was that his profile was not one where the United States risked losing a person who would otherwise be unable to enter the country. He was already here, the argument went, and the labor certification process exists precisely to test whether U.S. workers might fill his role. The RFE asked for substantive evidence that the labor certification process would cause specific harm to the national interest, beyond general arguments about delay.

Three challenges, each one anchored in legitimate adjudicator concerns, each one designed (deliberately or not) to force the petitioner into producing evidence that did not exist in the original filing.

The petitioner had 87 days to respond.

7. The Response Strategy

Three Problems, Three Specific Answers

I have a line I often repeat to new associates, sometimes more than once a day: ‘You don’t win an RFE by arguing harder; you win by showing the officer something they didn’t know was there.’

The temptation in responding to a multi prong RFE is to fight everywhere with equal energy: argue every position over again, overwhelm the adjudicator with volume, throw the kitchen sink at every concern. Visa Architect's experience, refined across hundreds of cases, is that this approach fails. Adjudicators read multiple RFE responses every day. They have developed an almost reflexive resistance to walls of repetitive argument. The responses that succeed are precise, evidence led, and structurally clean.

The strategy for Dr. Chakraborty's response was built around three principles.

The first principle was to treat each challenge as a separate investigation requiring its own new evidence, not a rehash of the original filing.

The second principle was to produce, where possible, evidence the adjudicator did not realize existed. Letters, documents, data points that change the underlying record, not just the rhetorical framing.

The third principle was to concede nothing essential, but to acknowledge the legitimate parts of the adjudicator's concerns. An RFE response that treats every adjudicator question as misguided weakens the response. A response that engages seriously with the question, then answers it with new evidence, builds trust.

For the national importance challenge, the response was structured around transcending the customer by customer framing.

The team obtained two new categories of evidence. The first was documentation of inter state data sharing facilitated by Dr. Chakraborty's surveillance platform. Several state health agencies that used the system had entered into mutual data sharing arrangements that allowed cross jurisdictional pattern detection, particularly important during the recent respiratory virus seasons. The petition documented specific instances where signals detected in one state's data had informed public health responses in neighboring states. This was not regional work that happened to occur in multiple regions. It was integrated multi jurisdictional surveillance.

The second category was federal level integration. The team obtained letters from two officials at federal level public health agencies confirming that data outputs from Dr. Chakraborty's platform had been incorporated into federal level situational awareness products during specific incident responses. One letter, from a senior official, described the use of the platform's outputs during a multi state respiratory illness investigation that the federal agency had coordinated.

The response also reframed the broader argument. The "national importance" prong under Dhanasar does not require that an endeavor be of national scope in every transaction. It requires that the endeavor have national-level implications. A platform that aggregates data from 1,800 hospitals in fourteen states, that feeds into federal situational awareness, that informs cross-state response coordination, is not a regional product that happens to be replicated in multiple regions. It is national infrastructure built one agency at a time.

For the well-positioned challenge, the response was structured around individual attribution within team contributions.

This was the most labour-intensive part of the response. The team worked with Dr. Chakraborty's employer to produce a detailed technical attribution document. The document broke down the surveillance platform component by component. It identified which algorithmic frameworks, which statistical models, which methodological choices, originated with Dr. Chakraborty individually. It distinguished those contributions from the work of software engineers, data engineers, and product managers on the broader team.

The Chief Scientific Officer of the company provided a supplemental letter that addressed the attribution question directly. The letter described, in specific technical terms, the original methodological frameworks Dr. Chakraborty had conceived. It identified which features of the platform existed because of choices he had made and would not exist, or would exist in a different form, without him. It also detailed his role in the disparities module, which had been substantially his individual conception.

The response added two new independent letters from academic researchers who had reviewed Dr. Chakraborty's published methodological work and could speak to which technical contributions in the public facing literature were attributable to him as first or sole author. One of the letters came from a researcher at a major academic medical center who had implemented elements of Dr. Chakraborty's published Bayesian framework in their own surveillance research.

Updated metrics were also included. In the months between the original filing and the RFE response, two additional state health agencies had begun using the platform. Citation counts on his methodological papers had grown by approximately 18 percent. A new preprint, with him as first author, had begun circulating in the public health informatics community.

For the balancing factor challenge, the response was structured around concrete harm.

The general argument, that PERM delays are bad for public health, was insufficient. The response had to make the harm specific.

The team obtained letters from three state epidemiologists addressing the specific question of what would happen if Dr. Chakraborty's involvement in their surveillance work were interrupted by an immigration process delay. Each letter described, in terms specific to that agency's operations, what the agency would lose. One letter detailed a specific multi year project, funded under a federal cooperative agreement, that depended on Dr. Chakraborty's continuity of involvement and would face substantial setbacks if his role were disrupted.

The response also addressed the U.S. worker availability question head on. It included new documentation, drawn from published workforce studies and from an expert affidavit by a senior academic in public health informatics, on the specific shortage of personnel with the combined epidemiological, statistical, and data engineering skill profile required for advanced surveillance work. The expert affidavit cited multiple peer reviewed workforce analyses describing this shortage as a documented constraint on U.S. public health system modernization.

Finally, the response made an argument that adjudicators sometimes find more persuasive than direct claims of irreplaceability. It argued that even if a hypothetical U.S. worker existed who could fill Dr. Chakraborty's role, the transition cost of replacing him during ongoing surveillance operations would itself impose national interest harm. Continuity of analytical methods matters in disease surveillance, because changes in analytical methods can produce apparent shifts in disease patterns that are actually artifacts of methodological change. Replacing the architect of a surveillance system in the middle of its operation is not a neutral act. It is a disruption with measurable consequences.

The response totaled 412 pages, including all new exhibits and letters. The legal brief portion was 31 pages. Every new piece of evidence was cross referenced. Every argument was anchored to specific documentary support.

It was filed on a Friday afternoon, eleven days before the response deadline.

Then, again, the waiting.

Fourteen months of work compressed into a 412-page answer to three questions. There was nothing more to do.

8. The Approval, and What It Actually Means

The Notice, and the Quiet Behind It

Forty-six days after the RFE response was filed, the approval notice arrived.

The petition had been approved on all three prongs: national importance, well-positioned, balance. The decision letter was, as approval letters often are, terse. It cited the framework. It listed the conclusions. It moved on.

For Dr. Chakraborty, the practical effect was substantial. As an India born applicant, his EB-2 priority date was now locked in. He could begin his adjustment of status process when his date became current. He could not be removed from the country on the basis of his status, as long as he maintained his lawful presence. He could continue, without disruption, the work he had spent the previous decade preparing to do.

He read the notice three times. He sat very still for about a minute. Then he laughed, the kind of laugh that comes when you have been holding your breath for fourteen months and finally let it out.

The morning the approval came through, he sent a one line email to his attorney at Visa Architect: Thank you. I think I need a long walk.

There is a temptation, when telling these stories, to wrap them in a triumphal bow. The petition was approved. The system worked. The applicant wins. Justice, of a kind, prevails. The truth is more complicated.

Dr. Chakraborty's case worked because it was built carefully, fought carefully, and supported by evidence that took months to assemble. It worked because his attorneys understood that the National Interest Waiver, properly used, is a precision instrument. The Dhanasar framework is unforgiving when applied loosely. It rewards specificity, evidence, and strategic restraint, and it punishes generality.

The case also worked because Dr. Chakraborty himself spent dozens of hours, over the course of more than a year, gathering documentation, sitting for interviews, reviewing drafts, and connecting his attorneys with the state and federal officials who could speak to his work. The case was not won by paperwork alone. It was won by a partnership between an applicant willing to do the work and a legal team that knew where to direct that work.

For every Dr. Chakraborty, however, there are public health professionals whose contributions are even harder to document. Epidemiologists working on rare disease surveillance, where outbreaks are infrequent and the absence of outbreaks is the measure of success. Data scientists building infrastructure for tribal health agencies, where political dynamics make official agency endorsements complicated. Researchers working in maternal mortality, opioid surveillance, and environmental health justice, fields where the work matters profoundly and the recognition mechanisms have not yet caught up with the work.

The American immigration system, in its highest aspirations, exists to identify and retain the people the country most needs. The National Interest Waiver category, when it functions as designed, is one of the more thoughtful instruments in that system. It does not ask whether you have a Nobel Prize. It asks whether your work matters to the country, whether you can do that work, and whether the country is better off if it lets you do it without the standard administrative friction.

The category, in other words, was designed for people exactly like Dr. Chakraborty. People whose contributions are real, important, and not always easy to see in the form of a journal citation or a media profile.

What separates the cases that succeed from the cases that fail is, in the end, often the quality of the translation. The translation of work into evidence. Of contribution into documentation. Of impact into language that an adjudicator, working through a stack of files at the end of a long week, can read, recognize, and credit.

Good work alone does not save you. Good work that someone has bothered to translate, that does.

That is the work Visa Architect was built to do.

Dr. Chakraborty's surveillance system is still running. Tomorrow morning, somewhere in a state health department, an epidemiologist will open a dashboard, see a flag, and decide whether it warrants investigation. Most of the time, the flag will be a false alarm, the kind of statistical fluctuation that appears in any large data system. Sometimes, the flag will be the first signal of something that needs attention: a localized outbreak, an unusual pattern, a disease that is moving in a way that ordinary clinical reporting has not yet caught.

When that happens, the patient who gets the earlier diagnosis, the community that gets the timely advisory, the public health response that arrives a day or two ahead of when it would otherwise have arrived, none of them will know the name of the person who built the system that flagged it. They will not know about his ten publications, his Bayesian ensemble model, his three pronged Request for Evidence, or the 412 page response that kept him in the country.

They will only know that the warning came in time.

That is, in the end, what this category is for.