Three people moved on a post, a family emigration and a student visa. In all three files, the record read as evidence rather than as a CV.

Three moves side by side. All three entries are drawn from reporting and primary records already published on this site, not from any filing.
Question	Katalin Karikó (1985)	Jan Koum (1992)	Kalpana Chawla (1982)
The mechanism of the move	A post-doctoral position abroad, taken after the lab she worked in at Szeged lost its funding	A family emigration at 16 as the Soviet Union came apart	A scholarship and a student visa for a master's degree
Age at the move	30	16	20
What made it hard at the point of exit	Hungary permitted the post but limited takeable money to $100 per adult, and gave her nothing because she was the one taking the position	Nothing was recoverable by effort: no language advantage, no degree, and a father left behind who never followed	Professors who told her aerospace was not a subject for women, and a father whose permission she needed and who opposed the move
What the record later showed	A 2005 discovery that made mRNA vaccines possible, and the 2023 Nobel Prize awarded for it	A messaging company scaled to roughly 450 million users and sold for about $19bn	Computational fluid dynamics research at NASA's Ames Research Center and two shuttle flights, including as a mission specialist on Columbia
The distance between the work and the recognition	Eighteen years, from the 2005 paper to the 2023 prize	Twenty-two years, from the food-stamp queue to the signature on the acquisition contract	The recognition arrived after her death; the Congressional Space Medal of Honor was awarded posthumously in 2004
What the move actually did	Removed a funding constraint, so the work became possible somewhere else	Put the family inside the one geography where the industry he would enter was being built	Converted an Indian engineering degree into American graduate credentials in the discipline her single dominant employer was hiring for

The short version

The three moves had almost nothing in common procedurally. Karikó left Hungary in 1985 on a post. Koum left Ukraine in 1992 as a 16-year-old in a family emigration. Chawla left India in 1982 on a scholarship. A post, a family, a student visa. Three different legal instruments, three different decades, three different destinations inside the same country.

What they share is not the route. It is what was already on the record about each of them when the opportunity came, and how that record was structured. In all three cases the record was not a list of positions held. It was a set of facts that demonstrated something specific: a piece of work that others built on, a product that a large number of people used, a piece of research in exactly the discipline the employer was hiring for.

That is a distinction most people assembling their own documentation miss. A CV says where you were. Evidence says what changed because you were there. The three files here differ in field, era and nationality, and the same asymmetry runs through all of them: the thing that made the next step possible was never the job title. It was the demonstration attached to it.

There is a second pattern, and it is the one that is most often missing from accounts of successful migrants. The move did not create the capability. It removed a constraint. Karikó did not become a better scientist on the flight to Philadelphia. Koum did not gain the ability to write software when the family reached Mountain View. Chawla did not become an engineer in Texas; she had qualified as one in Chandigarh. Each move made a constraint disappear — funding, geography, credential recognition — and the work that followed was work that had already been possible in principle, somewhere the constraint did not apply.

And there is a third pattern which is harder to like. In all three cases the interval between the useful work and the reward for it was long, and in one case it never arrived during the person's life. Karikó's cited discovery sat in the literature for eighteen years before Stockholm noticed. Koum's arrival and his exit are twenty-two years apart. Chawla's Congressional Space Medal of Honor was awarded posthumously on 4 February 2004, a year after Columbia broke up over Texas. Anyone reading these three for a template should read that part carefully. The move is the part under your control. The timing is not.

What follows is a comparison of the three mechanisms, then a comparison of the three records, then a short section on what does not transfer. The last section is the actionable one: three questions that these files answer for themselves and that most files do not.

Three mechanisms, three different ways out

A post: Karikó left because her funding was withdrawn

In 1985 the research programme Karikó worked in at the Biological Research Centre in Szeged lost its funding. She was 30. The constraint was not political in the ordinary sense and it was not about her ability. It was that the money for the work had gone and there was no domestic mechanism to replace it. Hungary at the time permitted its citizens to take a position abroad but restricted the money they could take with them: she was allowed $100, her husband and two-and-a-half-year-old daughter $50 each, and she, as the person taking the post, was allowed nothing at all. The family sold a Soviet-built Lada for the equivalent of about £1,000 and hid it inside her daughter's teddy bear to carry it across the border. The mechanism was an employment route. Her problem in 1985 was narrow and structural, and leaving solved exactly that problem and no other: the next fifteen years were still declined grants and a demotion at the University of Pennsylvania in the mid-1990s. What the post bought her was a position from which the work was still possible.

A family emigration: Koum left because the country he was born in came apart

Jan Borysovych Koum was born on 24 February 1976 in Kyiv and grew up in Fastiv. In 1992, at 16, he left with his mother and his grandmother for Mountain View, California. His father stayed behind, intending to follow; he never did, and died in 1997. There was no employer, no scholarship and no selection process. A social support programme placed the family in a small two-bedroom apartment, the family relied on food stamps, his mother worked as a babysitter and he swept floors at a grocery store. He kept his old Soviet school notebooks and used them. The mechanism was a family move with no professional content whatsoever, and the thing it delivered was location: he arrived inside the one geography where the industry he would later enter was being built by people his own age. That is luck, and he has never described it as a plan. What he did with it — a public high school in a language he had to finish learning, San Jose State, then nearly nine years as an engineer at Yahoo — is the part that was not luck.

A student visa: Chawla left because a scholarship made the degree possible

Kalpana Chawla was born on 1 July 1962 in Karnal, Haryana, the youngest of four children, and graduated from Punjab Engineering College in Chandigarh in 1982 with a Bachelor of Engineering in aeronautical engineering. She had been advised by professors to take electrical engineering instead, on the view that aerospace was not a suitable subject for women. In 1982 she travelled to the United States for postgraduate study on a scholarship to the University of Texas at Arlington, and the account of the family discussion is that her father opposed it and she went anyway. The mechanism was a graduate degree, which is the most ordinary of the three and the most transferable. It turned an Indian bachelor's degree into an American master's and then, at the University of Colorado Boulder in 1988, into a doctorate titled "Computation of dynamics and control of unsteady vortical flows" — computational fluid dynamics, which was precisely the discipline NASA's Ames Research Center was hiring for when it took her on that same year.

What the record showed, in each case

Set the mechanisms aside. The three records have a shape in common, and it is a shape worth naming because it is the thing that a person assembling their own documentation can actually work on.

In each case the record carried a fact that somebody other than the applicant had acted on. Karikó's 2005 paper, published to almost no attention, was eventually used: the pseudouridine-for-uridine substitution she and Drew Weissman found — the modification that silenced the inflammatory response to synthetic mRNA while increasing protein output — became the foundation of the mRNA vaccines produced against COVID-19 from early 2020. Koum's product was not just used but used at a scale no one disputed: about 450 million users and 55 employees at the time of the acquisition. Chawla's research was not merely published but adopted into the work of a specific institution, and then into flight: she operated the robotic arm on STS-87 in 1997, logged 10.4 million miles in 252 orbits by the end of that mission, and went on to head the astronaut corps's Crew Systems and Habitability department.

That is the difference between evidence and a CV in one sentence. A CV records that you held the position. Evidence records that something happened because you held it.

The second thing all three records do is survive being read by someone who was not in the room. None of the three depends on an adjective. Karikó's file does not say she is brilliant; it says she published a specific finding in 2005 and that the finding is what the Nobel Committee cited in 2023. Koum's file does not say he is a visionary founder; it says a company he co-founded had roughly 450 million users and was acquired in a deal announced on 19 February 2014 at approximately $19bn, comprising about $12bn in stock, $4bn in cash and $3bn of restricted stock vesting over four years. Chawla's file does not say she was dedicated; it says she held commercial pilot licences for land and seaplanes and gliders alongside a research post and a vice presidency at Overset Methods, Inc.

The third thing is corroboration. In each case the record includes someone else's account of the work, not only the person's own. Sequoia Capital's Jim Goetz wrote in 2014 that Koum's childhood had made him appreciate communication that was not bugged or taped. The Nobel Committee wrote the citation. NASA wrote the mission records and, later, the accident report. None of those are endorsements written at the applicant's request for the applicant's file. They are by-products of the work existing.

The mechanism is the least transferable part

It would be easy to read the three stories as a set of instructions about how to leave a country: get a post, get a scholarship, or have a family that emigrates. That reading is wrong in one specific way. Two of the three mechanisms were not chosen. Koum was 16 and the Soviet Union was dissolving; the decision was made by adults, in a hurry, under conditions nobody was optimising. Chawla's scholarship existed, but what she had to do was win it, persuade a father who opposed the move, and convert the degree afterwards into something an American research institution recognised as a credential in its own discipline.

The only mechanism in the three that resembles a deliberate career instrument is Karikó's post, and even that was a response to a funding decision she did not control. She described the period afterwards without decoration: people judged her unsuccessful when she felt very successful. The move did not make her judgement better or her research programme more likely to succeed. It made the work continue.

The transferable part is downstream of the mechanism. In each case, the move was followed by a period of converting what the person already had into the form the destination accepted. Chawla had to turn a Chandigarh bachelor's degree into a Texas master's and a Colorado doctorate in the exact discipline of her one dominant employer, then meet a legal condition — she became a naturalised US citizen in the early 1990s because citizenship was a requirement for the astronaut corps. Koum had to turn a Ukrainian school career into employable code, then spend nearly nine years as an engineer at Yahoo in someone else's codebase before the thing he built could exist. Karikó had to spend fifteen years being rejected in a funded system rather than an unfunded one.

None of that is a mechanism. All of it is a record being converted into a form that a stranger will accept.

What does not transfer

Each of the three is an extreme outlier and it would be dishonest to write them up as a method. The distance between these outcomes and an ordinary one is enormous. Karikó won a Nobel Prize. Koum signed a contract reported at approximately $19bn in an abandoned Mountain View building where he had collected food stamps as a teenager. Chawla flew on Columbia twice, and on 1 February 2003 the vehicle broke up above Texas and Louisiana sixteen minutes before it was due to land at the Kennedy Space Center; the Columbia Accident Investigation Board found that insulating foam shed from the external tank during launch had damaged the left wing and that the organisation's programme management had not treated the recurring foam loss as a safety-of-flight issue. All seven crew members died.

Two of the three stories contain a cost that no amount of planning addresses. Koum's father never left Ukraine and died in 1997; his mother was diagnosed with cancer and died in 2000, eight years after the move and nine years before WhatsApp shipped. Neither lived to see what the move made possible. Chawla's death is part of the record and belongs in any honest reading of it.

What transfers from all three is smaller and more durable than the outcomes. It is the structure of the constraint — something that could not be moved, in a place where it applied without exception — and the observation that once the constraint was gone, the next fifteen years were still hard. Karikó's story is not a story about recognition. It is a story about continuing to work on the same problem in a system that had stopped paying for it. Most people who move on that logic never get a prize. They get to keep working, which is the thing the constraint was preventing.

Three questions these files answer for themselves

What would have failed without you in it

Every one of these records contains a sentence of that shape. The modified-mRNA platform does not exist without the 2005 finding. The product does not exist without a co-founder who had spent nine years inside an engineering organisation. The research programme at Ames has her name attached to specific simulation work on airflow around aircraft such as the Harrier, with particular attention to ground effect. If your documentation cannot produce that sentence — not the job title, the counterfactual — you have a CV and not a record. This is the first and most common gap: an exhibit that is responsive to a subject and probative of nothing.

What did other people in the field do in response

Koum's corroboration is a Sequoia partner writing publicly, unprompted, about his childhood. Karikó's is a citation from a Nobel Committee and, before that, an industry that licensed and built on the modification. Chawla's is a set of mission records and a posthumous Congressional Space Medal of Honor awarded on 4 February 2004. In each case the proof of significance is an action taken by someone else, recorded somewhere that the applicant did not write. Testimonials that restate a definition carry no weight. Letters that describe what the person did, in the person's own words, transferred into the third person, carry no weight either.

What does the destination already accept as the credential

Chawla's move worked because computational fluid dynamics was the discipline NASA was hiring, and because she held American graduate degrees in it. Koum's worked because a dropout with strong code could get hired in the Bay Area in the 1990s. Karikó's worked because a funded research system somewhere else would pay for mRNA work even while it rejected her grant applications. The destination's acceptance criteria are not a detail to be explained in a covering letter. They are the specification the record has to meet, and if the record does not meet it, the move is the smaller half of the problem.

Questions this raises

Did Karikó, Koum and Chawla all move the same way?

No, and that is the point of the comparison. Karikó moved on a post-doctoral position in 1985 after her lab in Szeged lost its funding, Koum moved as a 16-year-old in a family emigration in 1992, and Chawla moved in 1982 on a scholarship and a student visa for a master's degree. Three different legal instruments across three decades.

What did the three records have in common?

Each carried a fact that somebody other than the applicant had acted on: a 2005 finding that became the foundation of mRNA vaccines, a messaging product with roughly 450 million users sold for about $19bn, and computational fluid dynamics research adopted into NASA's flight programme. In each case the record demonstrated something rather than listing positions held.

How long did each wait between the useful work and the recognition?

Karikó's cited discovery was published in 2005 and the Nobel Prize came in 2023, eighteen years later. Koum arrived on food stamps in 1992 and signed the acquisition contract in 2014, twenty-two years apart. Chawla's recognition arrived after her death, with the Congressional Space Medal of Honor awarded posthumously on 4 February 2004.

Is any of the three a typical case?

No. All three are extreme outliers, and two contain costs that planning does not address: Koum's parents both died before WhatsApp shipped, and Chawla died when Columbia broke up during re-entry on 1 February 2003. What transfers is the structure of the constraint, not the scale of the outcome.

What should someone take from this when assembling their own record?

Three tests. Name what would have failed without you in it. Show what other people in the field did in response to your work, in a source you did not write. And check what the destination already accepts as the credential — the degree, the licence, the citizenship, the discipline — before spending effort on anything else.