
| Country or region | Roles open there | Companies | Roles that may be done remotely |
|---|---|---|---|
| United States | 2,645 | 139 | 21% |
| Anywhere outside the United States | 1,164 | 97 | 31% |
| Canada | 401 | 52 | 49% |
| United Kingdom | 257 | 44 | 32% |
| India | 182 | 32 | 13% |
| Ireland | 94 | 20 | 40% |
| Germany | 92 | 19 | 54% |
| Spain | 73 | 14 | 75% |
| Poland | 69 | 17 | 54% |
| Latin America, six countries together | 34 | 13 | 26% |
Read on 29 September 2026 from the public Greenhouse, Ashby and Lever job boards. 3,441 roles with at least one country named; a role open in several countries counts in each row, and 263 of the Canadian roles are also open in the United States. Every role, every location string and the source for each company's home country are published in the full dataset.
One software engineering role in three at American tech companies is open outside the United States. One in a hundred is open in Latin America.
Those two numbers come from the public job boards of 148 US-based technology companies, read on 29 September 2026. We kept every software engineering role that names at least one country, 3,441 of them, and recorded every place each board says the role is open. Every role, every location string and every company is published, so the counts below can be checked row by row.
We went looking because the usual answer comes from vendor lists: Latin America for the time zone, Eastern Europe for depth, India for scale. Match.dev is one of those vendors. The companies’ own postings give a different picture.
1,164 of the 3,441 roles, 34 percent, are open in at least one country other than the United States. Most of those are not open in the United States at all: 796 roles, 23 percent of the total, name only places abroad. The other 368 can be filled on either side of the border.
Hiring abroad is common and also concentrated. 97 of the 148 companies have at least one engineering role open abroad, and ten companies supply 42 percent of those roles, led by Databricks, Okta, Grafana Labs, Affirm and GitLab. Without those five the share is 28 percent. Among the 78 companies with ten or more roles, the median company has 27 percent of them open outside the US, 23 companies have half or more, and 16 have none. The largest of the 16 is SpaceX, with 219 roles, all in the United States.
Of the roles open only outside the US, 62 percent are titled senior, staff, principal or an equivalent level. Of the roles open only in the US, 57 percent are. Titles are not levels, so the most this shows is that the roles abroad are not labeled junior.
Counted by every role open there, Canada is first: 401 roles at 52 companies, about one company in three. The United Kingdom follows with 257 roles at 44 companies, then India with 182 at 32, Ireland with 94 at 20, Germany with 92 at 19, Spain with 73 at 14 and Poland with 69 at 17.
That order depends on how you count, and the difference says something. 263 of the 401 Canadian roles are also open in the United States: the posting lists both countries. Count only roles that are not open in the US and the order is the United Kingdom with 177, India with 172, Canada with 138. Count only roles open in one country and nowhere else, and India leads with 165, ahead of Canada with 124 and the United Kingdom with 98.
Canada is where the most companies post roles, and India has the most roles posted for one country only. Spain is at the other end: 73 roles, 9 of them open in Spain alone. It is mostly one line in a list of countries.
Added up by region, Europe is the largest destination, with 506 roles at 70 companies. The rest of Asia-Pacific, mostly Singapore, Australia and Japan, adds 82 to India’s 182.
All of Latin America has 34 roles at 13 companies. That is 1 percent of the engineering roles in the sample and 3 percent of the ones open abroad. Brazil has 19, Mexico 9, Argentina 6, Colombia 2, Costa Rica and Uruguay one each. Poland alone has 69. Serbia has 34.
Lyft hires in Mexico City, two of its five roles there being internships. Braze, Brex and OpenAI hire in São Paulo, Coinbase remotely across Brazil, Webflow remotely in Argentina. In Brazil 8 of the 19 roles are titled senior or above, a lower share than in the US, on a sample too small to say more. The whole list fits in one table.
The count is a fact about job boards. Why it is so low is our reading, and we did not measure it. We treated every posting as an offer of employment. Employing someone in a country takes a legal entity there or an employer of record, and a company usually sets one up after it has decided to build a team in that place. By that reading the boards show where American tech companies have decided to build teams: Toronto, London, Dublin, Bengaluru.
The boards cannot see contractors, agencies or staffing platforms, which is the channel the vendor lists sell, ours included. This dataset says nothing about how large that channel is. It supports one narrow statement: in this sample, companies rarely post an engineering job in Latin America on their own board.
The usual argument for nearshoring is the shared workday. We measured it as the hours in which nine to five in the engineer’s city coincides with nine to five in New York, on the day of the read. Engineers who work with US teams often shift their day; this is the unshifted one.
Toronto shares all eight hours. São Paulo, Buenos Aires and Bogotá share seven, Mexico City six. London, Dublin and Lisbon share three. Berlin, Madrid and Warsaw share two. Bengaluru and Singapore share none. In winter, when US clocks move and most of Latin America’s do not, the Latin American figures shift by an hour in either direction.
From San Francisco the order changes. Mexico City shares seven hours, Bogotá six, Toronto five, São Paulo four, and every European city zero. Vancouver shares eight; 29 of the Canadian roles name it or British Columbia.
Each role open abroad goes into the group of its best-placed country. 429 roles, 37 percent, are in a country that shares six to eight hours with New York. 401 of those are open in Canada; 28 are same-clock roles without it. 26 percent share three hours, 14 percent share one or two, 19 percent share none, and the last 4 percent are in countries we did not put on the clock.
In these postings the same-clock role is a Canadian role. Latin America, the region that nearshoring vendors sell on the shared clock, accounts for 28 of the 429; five more Latin American roles are also open in Canada, and one is in Costa Rica, which we did not put on the clock.
“Hiring abroad” covers two arrangements, and the boards tell them apart.
In India, 13 percent of the roles are marked remote. Four in five of the rest name Bengaluru. 165 of the 182 Indian roles are open in India and nowhere else, which we read as a team that sits there and a role that belongs to it. 82 percent of the Indian roles are titled senior or above, and 78 percent with the largest employer, Databricks, set aside. Singapore looks the same, at 10 percent remote.
Spain is the opposite. Three roles in four are remote, and 64 of the 73 are open in other countries too. The most common list is Germany, Ireland, Spain, Sweden and the United Kingdom. Our reading: the company is hiring a person in a time zone, and Spain is one of the countries where it is able to employ one. Canada is in between, with about half its roles remote.
Job postings must carry a pay range in British Columbia since November 2023 and in Ontario since January 2026, and some companies print two bands in one posting. 74 roles at 19 companies give a base pay band for the United States and another for Canada, for the same job. 60 of those roles are at 17 US-based companies; the rest are at 1Password, which is Canadian, and Elastic, which names no headquarters.
Mercury’s senior backend engineer sits at the median. The US band is $166,600 to $208,300. The Canadian band is 157,400 to 196,800 Canadian dollars. The numbers are close and the currencies are different. At the European Central Bank’s reference rate for 28 September, 1.4166 Canadian dollars to the US dollar, the Canadian band is worth $111,000 to $139,000.
Across the 74 roles the Canadian number is a median 94 percent of the US number, and in 20 roles it is equal or higher. Brex prints identical figures for both countries, in different currencies. Converted to US dollars, the Canadian midpoint is a median 67 percent of the US one, and below three quarters in 68 of the 74 roles. For the 60 roles at US-based companies alone the median is the same, 67 percent. By company it runs from 57 percent at Samsara to 81 percent at Braze.
Two choices in that figure lean toward a smaller gap. 23 of the roles print several US bands for different pay zones, and we compared Canada with the lowest. Against the highest US band, the one a San Francisco or New York team pays, the median is 62.5 percent. And it is base pay only: equity, benefits and payroll taxes are not in these bands.
Bands are set in local currency, so the converted gap moves with the exchange rate. At 1.30 Canadian dollars to the US dollar the same bands would put Canada at 73 percent.
For Europe there are too few postings to give a figure. Eleven Elastic roles print a band in euros, at a median 73 percent of Elastic’s US midpoint, and one Notion role does, at 53 percent. Four roles print a band in pounds, at Klaviyo, Waymo and xAI, between 75 and 79 percent. Our reading of remote and office bands inside the US is a different question with a larger sample.
If you plan to employ engineers abroad, look at where companies like yours already do. Nine countries each have more than a dozen US tech companies posting engineering roles: Canada, the United Kingdom, India, Ireland, Germany, Poland, Singapore, Spain and Australia. In any of them there are people to ask, and for Canada there are published bands to compare against. We did not measure how hard it is to set up anywhere else.
Decide whether you are opening an office or hiring a person. A team in Bengaluru is a site: it needs a lead, enough people to stand on its own and work that belongs to it. A senior engineer in Valencia needs an entity or an employer of record, and no office.
Price the clock separately from the rate. Canada gives a New York team eight shared hours and a San Francisco team five to eight, at a base band about a third lower in US dollars. Europe gives New York two or three hours and San Francisco none. Teams that work by written handoff can absorb that; teams that decide things in conversation will feel it every day.
In Latin America, these companies rarely post engineering jobs on their own boards. The routes are an employer of record or a contractor, and this dataset cannot tell you which is more common. We found no posting that prints a US band next to a Latin American one, so there is no published number to anchor on, and a contractor arrangement carries a misclassification risk that is worth a lawyer’s hour before you sign.
The sample started as a hand-made list of technology companies, mostly venture-backed software companies and the public companies that grew out of them. We tried 387 board names on Greenhouse, Ashby and Lever. 189 names had a board that answered, two were dropped because the board under that name belongs to a different company, and 174 of the remaining 187 carried engineering roles. The largest employers run their own systems, so there is no Google, Microsoft, Amazon or Meta here. “American tech companies” in this article means this sample.
148 of the 174 are counted as based in the United States: 131 state a headquarters there, and 17 say they have no headquarters and were counted as US-based from a filing, a mailing address or a third-party listing, among them Block, Coinbase and GitLab. The basis for each is in the table. With only the 131, the share of roles open abroad is 33 percent. The other 26 companies are based elsewhere, name headquarters in two countries, or state no home we could source. With the last two groups added back the share is 36 percent, and the first four countries are the same in the same order.
Postings were merged into roles. Some boards publish one role as a separate posting for each place, so postings of one company with the same title and the same opening 400 characters count as one role. Counted in postings instead, the share open abroad is 37 percent. A role open in five countries is counted in all five, though it is usually one hire, so country rows add up to more than the total.
A role is an opening, and some openings are permanent. 363 of the 3,441 were published more than a year before the read date, and may be standing postings. Region words were ignored: 42 roles counted as open in the United States only also say “North America”, “Americas” or “EMEA”, which undercounts Canada and the abroad share a little. We did not check the type of contract behind a posting.
Countries were read from the boards’ location fields by written rules. A language model then read all 882 distinct location strings, and a second model, given the place strings of 100 random roles and nothing else, agreed on all 100. No person checked these rows, which is why all of them are printed. Where each company is based was looked up company by company, and the source is in the table, including the cases where the only source is Wikipedia or a press dateline.
All of it, every role with its places, its countries and its link, and every pay band compared here with the sentence it came from, is published.
Match.dev places senior remote engineers, most of them in Latin America and Eastern Europe, as contractors. Clients pay $50 to $80 an hour, and every engineer has passed a paid trial on a real task. That is the channel this dataset cannot see, so read the section on Latin America as written from inside it. If you want to see candidates, tell us what you are building.
This website uses its own and third-party cookies for analytical purposes and to offer you advertising of interest to us and third parties. You can consult all the information in our Cookies Policy. You can manage the acceptance or rejection of cookies by clicking on “Configure”.
Privacy is important to us, so you have the option of disabling certain types of storage that may not be necessary for the basic functioning of the website. Blocking categories may impact your experience on the website. More information


