Level 5 Data Engineer Apprenticeship (ST1386) UK

By James Cotton, Founder, iO-Sphere

Most roles put forward for this apprenticeship are not data-engineering roles. That is the assumption every other guide leaves unexamined: it takes the role in front of you as given and moves straight to levels and funding.

Often the role is something else. A data engineer builds the systems that collect, move, clean and store data so that analysts, dashboards and AI tools have something reliable to work with. An analyst uses what those systems produce. The two share tools and vocabulary, which is why they get mixed up, but they are separate occupations. Level 5 here certifies a different job from the Level 4 Data Analyst route rather than a more advanced version of it.

So a guide that hands you a level description, an entry-requirements list and no way to decide leaves out the one question that governs the rest. Choose the highest level a person could be accepted onto, rather than the level the job actually contains, and you fund a syllabus the day job cannot feed.

Is the role really data engineering?

The Data Engineer standard, ST1386, is the national specification for what a data engineer must be able to do to qualify. It is maintained by Skills England, the government body that approves apprenticeship standards. The standard describes production engineering work: building and maintaining data pipelines, standing up and running the platforms that store and process data, writing and looking after transformation logic in Python and SQL, and doing all of it to the reliability that live systems demand.

Key figures at a glance

Standard and level: Data Engineer (ST1386), Level 5
Off-the-job hours: 418, set by the standard
Funding band: Up to £19,000 (Skills England)
Typical duration: 24 months; our delivery is 15 months of training plus a 3-month end-point assessment
End-point assessment: A project report and a professional discussion, with no written exam
Entry requirements: Set by the training provider, not the standard

Here is the test we apply, and the one we would ask you to apply before anyone enrols. A role holds Level 5 data-engineering work if it involves at least two of these:

owning a production data pipeline end to end;
choosing and running a data storage or processing platform;
writing and maintaining transformation logic in Python or SQL at a scale where simple approaches break;
guiding others on how the data infrastructure should be built.

It does not fit if the role's outputs are reports, dashboards or one-off queries, however advanced. Those sit downstream of the engineering: they consume the pipelines rather than build them.

The mix-up we see most often is a capable analyst taken for an engineer, because both write code and both work with data. The quickest way to tell them apart costs nothing: ask the line manager to list the pipeline-building tasks the role will actually be assigned. If that list keeps turning into questions they want answered from the data, the role is analysis, and Level 5 is the wrong certificate for it.

The hours behind the test

There is a number attached to that work, and it is why the question cannot be waved through. Every apprenticeship standard now sets its own minimum of off-the-job training hours; since August 2025 that figure is published per standard, replacing the old "twenty per cent of contracted hours" proxy. Off-the-job hours are not classroom time and they are not the apprentice's own time: they are protected, paid hours inside the normal working week, spent on development the person cannot yet do. ST1386 sets 418 of them, the most of any standard we deliver.

Because they are protected working hours spent doing Level 5 engineering, they have to come out of a real job. We smooth the standard's figure across delivery, so over our 15 months of training the 418 hours work out at roughly six and a half hours a week. In practice, then, the role has to contain about six and a half hours a week of genuine pipeline-building for someone to fill the requirement. If the line manager cannot point to that work, no syllabus and no funding will create it.

If the work isn't there, here is where to go instead

If the role does not contain data-engineering work, no pipelines to build, no platform to run, no data moving at scale, then Level 5 is the wrong choice however capable or keen the person is, for the reason the summary above gives.

Three routes are open instead, and it helps to name where each leads rather than gesture at a lower level:

The role produces reports, dashboards or analytical findings. The fit is the Level 4 Data Analyst standard (ST0118), which develops people who work with existing data to answer business questions; it carries 370 off-the-job hours. Ours is the Level 4 Data Analyst apprenticeship, and if the choice between the two jobs is the whole question, our data engineer vs data analyst comparison draws the line in detail.
The work is foundational data handling, closer to preparing and looking after data than to building the systems that move it. The fit is Level 3, the Data Technician standard (ST0795), at 348 hours. Ours is Data & AI Essentials, and the Level 3 route explained in full covers who it suits.
Nobody in the role does data engineering yet. Then no apprenticeship is the right answer this year: build the work first, or start the person at the level the job holds now and revisit later. The employer guide walks through both.

And if the ambition is statistical modelling and machine learning rather than the pipelines that feed them, that is data science. It is a separate discipline that iO-Sphere does not deliver, so it is not a route we can place you on here.

If the work is there, here is what you are committing to

Everything below assumes the role contains the work. If it does not, none of it changes the answer.

The standard, and where Level 5 sits

As a qualification level, Level 5 sits above A-level and below a full bachelor's degree. That describes how demanding the work is. It says nothing about how senior the person is, or how a data-analyst role ranks against a data-engineer one. Choosing between Level 4 and Level 5 is a question about which job the role actually is, and answering it that way is what keeps the decision sound.

One fact matters if you are timing a start: ST1386 is approved for delivery now, and a revised version of the standard applies to new starts from 18 December 2026, so if you are planning a start around then, check the live standard page for the version you will be on. Two bodies sit behind the route, and it helps to keep them straight: Skills England approves the standard and sets the £19,000 funding band, while the Department for Work and Pensions now owns apprenticeship funding policy and the funding rules.

What the apprentice needs before starting

Entry requirements are set by the training provider, not by the standard: Skills England publishes none for ST1386, so they vary from provider to provider. Our own typical requirement is a Level 3 qualification in a relevant area at any grade (two A levels in a related subject, a Level 3 apprenticeship, or a BTEC Extended Diploma, a vocational qualification broadly equivalent to three A levels), or genuine technical grounding in its place, such as working knowledge of Python or two to three years in a related role. Ask any provider for theirs.

This route needs no degree, but it does need prior knowledge, because Level 5 is engineering work and it assumes a real technical starting point.

Apprentices also need English and maths to Level 2, roughly a GCSE pass at grade 9 to 4, or an Ofqual-regulated Functional Skills qualification at Level 2. For anyone aged 19 or over at the start, this is an optional exit requirement, meaning they can finish the apprenticeship without it, and it has been since 11 February 2025, though it can still be funded where they lack it and the employer agrees; for 16 to 18-year-olds it remains a condition of completing.

One group is often the strongest fit of all: existing employees. Someone who already knows your systems, formalising and deepening work they can already partly do, is frequently the best use of the route, as long as the role carries the engineering work.

What it covers, and how it is assessed

The standard's content is production data engineering from end to end: designing and building pipelines, working in Python and SQL, managing how data is stored and processed, and applying data governance and security so the result is reliable enough to trust. We teach it on real data and real problems rather than tidy examples. Where a workplace cannot practically provide a particular scenario, apprentices first work through it on realistic business case studies and simulations (our Prism method) before applying it to your own systems.

Assessment is where a work-based route parts company with a classroom course. There is no written exam. The end-point assessment (EPA), the independent check at the end of an apprenticeship, is built from two pieces of the apprentice's own work: a project evaluation report, which they present and answer questions on, and a professional discussion. Both are graded, and strong work earns a distinction.

It is carried out by an independent assessment organisation that the employer chooses from the official register, separate from whoever delivered the training. For an apprentice who finds written exams stressful, that changes the experience: they are judged on what they can build and explain rather than on what they can recall against the clock. Our end-point assessment glossary entry sets out the mechanics.

How long it takes

iO-Sphere delivers the Level 5 Data Engineer apprenticeship in 15 months of training plus a three-month end-point assessment. The standard's typical duration is 24 months, and providers arrange delivery differently within that: Cambridge Spark, for example, advertises 14 months of training. Duration reflects how a provider has structured delivery; the standard's demands are the same whoever teaches it. Throughout, the off-the-job hours stay protected, paid working time.

What it costs

A start you plan now falls in the 2026-27 funding year, which runs from 1 August 2026. The funding band is the maximum the government will put toward training and assessment: for ST1386 it is £19,000, the band set by Skills England. Here is who pays what within it.

If you pay the apprenticeship levy (an annual pay bill above £3 million), you draw on your apprenticeship service account, which covers training and the end-point assessment up to the £19,000 band.

If you do not pay the levy (a pay bill under £3 million), the apprenticeship is still funded. The government meets the cost in full for an apprentice aged 16 to 24 at the start. For an apprentice aged 25 or over, it meets 95 per cent and you contribute the remaining 5 per cent, an arrangement called co-investment. Alternatively, a larger employer can transfer levy funds to cover up to the full cost.

If a levy account runs out mid-year, the 2026-27 rules still fund a 16 to 24-year-old at the start in full; for an apprentice aged 25 or over the split becomes 75 per cent government and 25 per cent employer.

One aside, because it comes up: the Level 7 funding withdrawal, which stopped levy funding for master's-level apprenticeships from January 2026 and only for starts aged 22 and over, does not touch Level 5. The full levy mechanics, and the case that those 418 protected hours are new capability the levy is designed to buy, sit in our Growth and Skills Levy guide for data engineering. You can confirm the live band in the apprenticeship funding rules before you commit budget, since bands are reviewed periodically.

One cost sits outside the funding band: the apprentice's wage, which you pay. The legal minimum apprentice rate is £8.00 an hour from 1 April 2026. That apprentice rate applies only while the apprentice is under 19, or in the first year of the apprenticeship if they are 19 or over; after that they move to the ordinary minimum wage for their age. It is a legal floor, and most employers pay data-engineering apprentices well above it.

Where it leads

A completed Level 5 Data Engineer apprenticeship qualifies someone for data engineer roles and gives them a strong base to grow from, toward senior engineering and, in time, machine-learning engineering, which has its own Level 6 standard (ST1398) that iO-Sphere does not deliver. Data engineering skills are in short supply and the work is production-critical, which tends to make the people who hold these roles worth keeping.

That same scarcity shows up in the pay. Advertised medians skew high, since they reflect experienced hires, and sit well above what an entry-level engineer earns in the first few years. There is no data-engineer-specific ONS occupation code, so no clean measured national figure exists; for a realistic number, take a current, dated figure from a recruitment salary guide for your own stack and region rather than trusting a single headline.

There is also no outcome rate for this apprenticeship yet, because our apprenticeship line is new and an honest completion or progression figure does not exist to quote.

What to ask us, and everyone else on your list

Whether you talk to us or to anyone else, sort providers by four questions, in this order:

What will the learner actually do? If the answer is "watch lectures", stop there. If it is "build pipelines on real data", keep going.
Who coaches? People who have done the job, or people who teach the subject? Ask outright, and ask to see backgrounds.
Does the role contain Level 5 work? Confirm with the line manager which tasks they will assign. This is the check most shortlists never make.
What does it run on? It should be ST1386, at Level 5. Fit the funding to that, last.

Funding comes last on purpose. "Is it free?" is the wrong first question. Decide the outcome first (what you want this person to be able to do, and how you will know they can), then judge every route as if you were paying for it in cash, and only then work out how it is funded.

Two of those answers are ours to be held to. We coach with practitioners who have done the job, and you can ask about any of them. And we will not enrol someone whose role does not hold Level 5 engineering work. Hold every provider on your list to the same two.

FAQ

What level is the Data Engineer apprenticeship?

It is Level 5, delivered against the national Data Engineer standard, ST1386. Level 5 describes the complexity of the work, sitting above A-level and below a full bachelor's degree; it does not describe seniority. The certificate is for a data engineer, which is a different occupation from the Level 4 data analyst rather than a more advanced version of it.

How long does the Level 5 Data Engineer apprenticeship take?

The standard's typical duration is 24 months. iO-Sphere runs it as 15 months of training followed by a three-month end-point assessment, and other providers structure delivery differently, so advertised lengths vary. The figure worth weighing is how much protected engineering time the programme actually contains, more than the headline length.

What are the entry requirements?

Skills England sets no entry requirements for this standard, so each provider defines its own. Ours typically look for a Level 3 qualification in a relevant area at any grade, or comparable technical grounding such as working Python or a couple of years in a related role; ask any provider for theirs. There is no degree requirement, but there is a real requirement for prior technical knowledge. Separately, apprentices need English and maths to Level 2, which since 11 February 2025 is an optional exit requirement for anyone aged 19 or over at the start, and a completion condition for those aged 16 to 18.

How is the Level 5 Data Engineer apprenticeship funded?

The funding band for ST1386 is up to £19,000, set by Skills England and drawn through the apprenticeship levy. Who pays what within it depends on your pay bill and the apprentice's age; the "What it costs" section above sets out the levy, non-levy and exhausted-pot cases in full. The short version: an apprentice aged 16 to 24 at the start is fully funded for every type of employer, while for those aged 25 or over the employer contributes either 5 per cent under non-levy co-investment or 25 per cent if a levy account is exhausted.

Is there an exam?

There is no written exam. The end-point assessment is built from two pieces of the apprentice's real work: a project evaluation report they present and take questions on, and a professional discussion, both marked by an independent assessment organisation the employer picks from the official register. The point of the format is to judge someone on what they can build and explain in their own work.

Should I pick Level 5, or a lower level?

Pick the level whose work the role actually contains now. Level 5 fits when the role actually does at least two of: owning a production pipeline end to end, running a storage or processing platform, maintaining Python or SQL transformation logic at real scale, or guiding others on data infrastructure. If the role's output is reports, dashboards and answers to business questions, that points to the Level 4 Data Analyst standard (ST0118); if it is foundational data handling, to the Level 3 Data Technician standard (ST0795).

Can an existing employee do the Level 5 Data Engineer apprenticeship?

Yes, and it is often the best use of the route. Where someone already does part of the engineering work but lacks the formal grounding, the apprenticeship builds that grounding on the systems they already know. The eligibility conditions match those for a new hire: a Level 3 qualification or equivalent experience, English and maths to Level 2 (an optional exit for those aged 19 or over since February 2025), and, above all, a role that genuinely contains Level 5 engineering work.

The next step

If that Level 5 work really lives in the role, the next step is a conversation about what your team will build, not a brochure. Look at our Level 5 Data Engineer apprenticeship, or talk to us about whether it is the right level for the work you have. And if it is not, the more useful outcome is that you now know which route is.

External figures on this page (the ST1386 funding band and assessment plan, and Cambridge Spark's advertised duration) were checked in August 2026; funding bands and provider details change, so confirm the live figure before you commit budget.