What 335 Remote Engineering Jobs Actually Require in 2026
Everyone has an opinion about what remote engineering hiring looks like right now. Fewer people have counted.
So we did. Every job on JobRemotely has to publish a salary before we list it — that is a hard filter, not a preference — which leaves us with an unusual dataset: 335 remote engineering roles, all live, all with a real compensation range in USD. No "competitive salary". No guessing.
Here is what those 335 job descriptions actually say.
How we did this
Being straight about method matters more than a clean headline, so:
- Dataset: all active engineering-category listings on JobRemotely on 28 July 2026 — 335 roles out of 667 total jobs.
- Salary: every listing has a disclosed range, normalised to annual USD at import. Where we quote a "median", it is the midpoint of the range.
- Technology counts: case-insensitive whole-word matches inside the job description. This matters — a naive substring search reports "Rust" in 41 % of ads because it matches the word t-rust-ed. Ours does not.
- What a count means: a technology is mentioned in the ad. That is not the same as "you will be rejected without it".
- Known skew: our sources cluster in two places — US companies and Polish employers. So this is an accurate picture of this board, not a global census. We flag geography wherever it changes the answer, because it changes it a lot.
Finding 1: AI stopped being a specialisation
The single clearest signal in the data.
| How we count it | Engineering roles | Share |
|---|---|---|
| Explicit — "machine learning", "LLM", "artificial intelligence", "GenAI" | 114 | 34.0 % |
| Loose — the above plus any standalone mention of "AI" | 204 | 60.9 % |
Take the conservative number: one in three remote engineering ads now spells out AI, ML or LLM work. Not in a dedicated "AI Engineer" title — inside ordinary backend, platform and product engineering roles. Across the whole board (all 667 jobs, every category) the strict figure is 36.9 %.
Two years ago that language lived in a handful of research postings. It is now closer to a default expectation, the way "familiar with cloud" became one.
Finding 2: mentioning AI tracks with a 54 % higher salary
We compared US-based roles only, to strip out the geography effect described below, and used the strict definition again (explicit ML / LLM / AI language, not a passing "AI-powered" in the company boilerplate).
| US remote roles | Count | Median salary |
|---|---|---|
| Mention AI / ML / LLM | 165 | $246,000 |
| Do not | 157 | $159,500 |
That is an $86,500 gap at the midpoint.
The honest caveat: this is correlation. Roles that mention AI skew senior, and they skew toward well-funded US tech companies that pay above market anyway. Learning a vector database will not add $80k to your offer by itself. What the number does show is where the budget is currently pointed — and that is worth knowing when you choose which of two similar roles to prepare for.
Finding 3: the cloud is not one market
If you assume AWS has run away with it, the ads disagree.
| Platform | Roles | Share |
|---|---|---|
| AWS | 58 | 17.3 % |
| Azure | 40 | 11.9 % |
| Google Cloud | 20 | 6.0 % |
AWS still leads, but Azure sits closer than its reputation among developers would suggest — largely because enterprise and Poland-based employers in our dataset lean Microsoft. If you have only ever touched one provider, the second one is cheap to learn and visibly widens the funnel.
Finding 4: the platform baseline
| Skill | Roles | Share |
|---|---|---|
| CI/CD | 57 | 17.0 % |
| Kubernetes | 45 | 13.4 % |
| Docker | 42 | 12.5 % |
| Terraform | 17 | 5.1 % |
Containers and pipelines now appear in ordinary product-engineering ads, not only in DevOps roles. Kubernetes at 13 % is the interesting one: high enough that it keeps showing up, low enough that "I have shipped to k8s" is still a differentiator rather than an assumption.
Finding 5: the language layer is unsurprising, and that is the point
| Language / framework | Roles | Share |
|---|---|---|
| React | 60 | 17.9 % |
| TypeScript | 57 | 17.0 % |
| SQL | 55 | 16.4 % |
| Python | 55 | 16.4 % |
| Java | 38 | 11.3 % |
| JavaScript | 35 | 10.4 % |
| PostgreSQL | 32 | 9.6 % |
| Node.js | 31 | 9.3 % |
| Kotlin | 21 | 6.3 % |
| Go | 18 | 5.4 % |
| GraphQL | 14 | 4.2 % |
| Rust | 6 | 1.8 % |
| Swift | 5 | 1.5 % |
React and TypeScript lead the front end. Python and SQL are the quiet constants — SQL turns up in a sixth of engineering ads, before you even open the data roles, where it is closer to table stakes.
Rust at 1.8 % deserves a note. It is a genuinely excellent language with a loud community and a small hiring market. Learn it because you want to; do not learn it because you think it is where the remote jobs are.
Finding 6: "remote" is not one salary market
This is the finding that recontextualises every number above.
| Location | Roles | Median salary |
|---|---|---|
| United States | 322 | $199,750 |
| Poland | 287 | $74,844 |
| United Kingdom | 5 | $157,400 |
| Canada | 15 | $129,158 |
| European Union | 9 | $122,575 |
Same word — "remote" — and a 2.7× difference at the median. Both sets of roles are fully remote. Both are real jobs at real companies. The difference is which market the employer benchmarks against.
A practical consequence: any headline that says "remote engineers earn X" is almost meaningless without a location filter. When you compare two offers, compare them inside the same benchmark, and when a company lists a wide band, ask which market the band is anchored to. That single question tells you more than most salary reports.
What we would take from this as a job seeker
- Put AI work on your CV if you have genuinely done it — even internal tooling or evaluation work. A third of ads now use that vocabulary, and screening reads for it.
- Know a second cloud. Azure is closer to AWS in these ads than most developers assume.
- Ship something to containers. CI/CD plus Docker plus a little Kubernetes covers the platform expectations in most of these roles.
- Read the salary before the title. "Senior" means two very different numbers depending on the market the role is benchmarked to.
- Filter for disclosed pay. Ads that publish a range attract far more applicants precisely because they respect your time — which is why we refuse to list roles without one.
The data behind this post
Every figure here comes from live listings you can open right now: remote engineering jobs, or the full board across every category. Each role is vetted, remote, in English, and shows its salary before you click apply.
We will re-run this analysis as the board grows, so the numbers stay honest rather than frozen at a moment that flattered us.