Live job-market report Data Engineer Updated today

Data Engineer Skills — What Job Posts Actually Ask For (2026)

Aggregated from live Data Engineer postings rather than opinion. Every figure below is measured from real adverts and recalculated daily.

10,967 live Data Engineer postings read
Hiring companies 4,114
Added this week 2,284
Posted in last 24h 111

What a Data Engineer actually does

A data engineer builds and runs the plumbing that moves data from wherever it is produced to wherever someone needs it — pulling from production databases, event streams and third-party APIs, reshaping what comes back, and landing it somewhere an analyst or a model can query without knowing where any of it came from. The work is closer to backend engineering than to data science, and most of it is really about failure: a pipeline that works on Tuesday and silently drops half its rows on Wednesday is worse than one that never ran, because the numbers downstream still look plausible.

The skills Data Engineer postings actually ask for

Aggregated from the live descriptions of postings whose title names the role. Not a shortlist, and not anybody's opinion of what matters.

  • python 72.3%
  • artificial intelligence 63.3%
  • sql 61.5%
  • pipeline 60.3%
  • data engineering 52.9%
  • analytics 52.4%
  • data pipeline 50.4%
  • infrastructure 50.3%
  • security 41.2%
  • data quality 41.1%
  • aws 39.6%
  • machine learning 36.1%
  • etl 35.4%
  • spark 33.7%
  • monitoring 33%
  • software engineer 32.3%

Share of Data Engineer postings mentioning the skill — a posting asks for many, so these do not add to 100. Measured across postings whose title is the role.

How much experience, and where you would work

Minimum experience asked for

  • 0-2 years 1,272 12%
  • 3-5 years 4,883 45%
  • 6+ years 2,953 27%
  • Not stated 1,859 17%

Work policy

  • Remote 3,484 32%
  • Hybrid 2,509 23%
  • On-site 3,166 29%
  • Not specified 1,808 16%

423 of those 3,484 remote Data Engineer roles are open to candidates anywhere in the world — the rest are remote within one country only. Most job boards do not separate the two.

What separates a senior posting from a mid-level one

The same skill list, split by the seniority of the posting asking for it. Bars run right where senior postings ask more often, left where mid-level postings do.

  • mentoring +22.9 pt
  • artificial intelligence +13.1 pt
  • spark +12.9 pt
  • data engineering +10.9 pt
  • pipeline +10.5 pt
  • power bi -3.3 pt
  • data warehouse -3.5 pt
  • problem solving -3.9 pt
  • etl -3.9 pt
  • version control -4.3 pt

Percentage-point gap between the senior+ and mid-level shares naming each skill. Senior+ includes principal, staff and executive.

The skills that are asked for rarely — and are worth more for it

Everyone asks for the top of that first chart, so naming it proves nothing. These sit far enough down to differentiate you, and close enough to the top to be worth the effort.

  • 14.5% terraform of live postings
  • 14.4% devops of live postings
  • 14.4% bigquery of live postings
  • 14.1% use cases of live postings
  • 13.9% big data of live postings
  • 13.7% code review of live postings
  • 13.5% infrastructure as code of live postings
  • 13.4% large language models of live postings
  • 13.1% scala of live postings
  • 12.6% docker of live postings

Who is hiring, and where those roles sit

Companies hiring for this role

  • 1 Nvidia 356
  • 2 Databricks 339
  • 3 Salesforce 167
  • 4 Intel 155
  • 5 Canonical 138
  • 6 Mastercard 124
  • 7 Esri 123
  • 8 Snowflake 118

Where the roles we index sit

Where our listings sit, not the world's — the corpus skews to US and English-language boards.

Latest openings

Updated daily · closed roles removed within 24h
See these roles ranked just for you Upload your CV once. Every role is scored 0–100 against your experience, in real time as you search. Rank jobs by my fit →
See all 10,967 live postings →

Free. No card required. Every role ranked against your CV.

Summary

The top of that chart is table stakes. SQL and Python appear in roughly four out of five postings, which means naming them on a CV distinguishes you from nobody — their absence is disqualifying and their presence is unremarkable. The useful reading starts in the middle of the list, where the split between the warehouse-and-orchestration stack and the cloud platform tells you what kind of team is hiring.

The rarer skills further down are where the leverage is. A skill appearing in one posting in eight still represents a substantial number of open roles, and it is specific enough that a CV carrying it gets read differently. That is the practical use of a measured list over a written one: an article cannot tell you which skills are rare, because nobody ever measured how common any of them were.

Where these roles are, and how to get to them

Data engineering is one of the more internationally mobile technical roles, largely because the work does not depend on local market knowledge. The country breakdown above links into the destinations that sponsor visas, and the remote share is worth reading carefully — a role listed as remote is usually remote within one country, which is a different proposition entirely if you are not already in it.

If you are weighing whether to apply, the fastest way to use this page is to compare it against your own CV rather than against the market. Upload it to JobGlance and every posting here is scored 0–100 against your experience, so the gap between what these adverts ask for and what you can evidence stops being a guess.

Common questions

On the evidence above, SQL and Python are near-universal, followed by pipeline and ETL work, a cloud platform, and data modelling. The exact figures and the full ranked list are at the top of this page, measured from live postings and recalculated daily rather than taken from a fixed list.

The largest band by some distance is three to five years, with a meaningful tail asking six or more and a noticeably smaller entry-level share. Read those as minimums: the figure is the lower bound a posting states, so an advert saying "5+ years" sits in the three-to-five band, not above it. A fifth of postings state no requirement at all.

The distribution above says the honest answer is that it is hard to enter and comparatively easy to stay in. Entry-level postings are the smallest experience band, and the seniority chart shows why: what employers pay more for is modelling and architecture judgement, which is difficult to demonstrate without having already done the job. The common route in is sideways, from analytics or backend engineering, rather than directly.

A substantial share of these postings are remote — the exact split is in the work-policy chart above, alongside the number that are open to candidates in any country rather than remote within one. That second figure is much smaller than the first, and it is the one that matters if you are not already living where the employer is.

Very few of these postings make a degree the headline requirement, and the skills they do lead with are all demonstrable without one. What is consistently asked for is evidence of having built and operated something that runs unattended, which a portfolio can carry and a transcript generally cannot.

Keep looking

Data analyst skillsAll role insightsScore your CV against these roles

See which of these you actually match

Upload your CV once. Every role here gets scored 0–100 against it, so you apply to the ones where you have a real shot.

Get my match scores → How Smart Match works

Free to start · no card required