Growth & Skills Levy: Data Engineering (Level 5)

By James Cotton, Founder, iO-Sphere

Most levy pages stop exactly where your decision starts. They explain the pots and the rules, then leave you at the counter. Here's what they skip: funding determines the route, not the result. A funded programme that doesn't change what someone can do is the most expensive training there is, because it spends the dearest resource you have — working time. So run this decision outcomes-first. Name the data engineering capability you need, decide how you'll see it in the work, then let the funding follow. Do that and the levy is genuinely generous. Lead with the funding and you can end up with a certificate and no changed work.

This page is for HR and L&D operators and employers who know their levy budget exists but haven't yet worked out how the Growth & Skills Levy changes eligibility, flexibility and provider choice versus the old Apprenticeship Levy.

Key figures at a glance

Levy rate

0.5% of an employer's annual pay bill (HMRC PAYE, unchanged since 2017)

Pay-bill threshold to pay the levy

£3 million annual pay bill; £15,000 annual allowance offsets the charge (unchanged since 2017)

Current name of the levy

Growth & Skills Levy — renamed and redesigned from the Apprenticeship Levy, April 2026

Maximum levy funds transferable to another employer

50% of unused funds, raised from 25% on 22 April 2024 (CIPP)

Data Engineer apprenticeship standard

Level 5, Data Engineer (ST1386) — a Skills England–approved standard (findatrainingprovider.co.uk, 2026-06-15)

Data engineer pay by experience

Entry-level ~£31,758; 1–4 years ~£42,853 average total compensation (reported — payscale.com, accessed 2026-07-01)

What is the Growth & Skills Levy?

The Growth & Skills Levy is the renamed, redesigned version of the Apprenticeship Levy, effective from April 2026. The underlying mechanic hasn't changed: employers with an annual pay bill over £3 million pay 0.5% of that pay bill, collected monthly through PAYE and offset by a £15,000 annual allowance. What changed is the name and the flexibility around what the money can buy. The plumbing is the same; the taps are being widened.

If your organisation pays the levy, the funds sit in your Digital Apprenticeship Service (DAS) account and can pay for approved apprenticeship training and its end-point assessment. They can't be spent on wages, travel, or other employment costs. The levy funds the training; you fund the person.

The most talked-about reform is flexibility. From April 2026 onward, phased in over time, up to 50% of the levy is intended to be usable on shorter, flexible non-apprenticeship training units — some as short as a week — rather than only on full apprenticeships. Treat the exact scope as still settling, and check the latest DfE/DWP funding rules before committing to a non-apprenticeship spend, because the mechanics are phasing in.

For clarity on where we stand: iO-Sphere doesn't deliver apprenticeship units — the funding attached is small and the format constrains depth. A unit fits when an employer needs a fully-funded, short, narrow intervention and can't fund more; it's not what we'd recommend where genuine capability-building is the goal.

How the Growth & Skills Levy differs from the Apprenticeship Levy

The core difference is flexibility, not the money itself. The Apprenticeship Levy could only fund approved apprenticeships and their assessment. The Growth & Skills Levy keeps that as the main use but adds the ability to spend a portion — up to 50%, phased from April 2026 — on shorter, non-apprenticeship training. That matters if you want to top up an existing team's skills without committing every learner to a full 15-month programme.

What stayed the same is worth stating plainly, because it's where plans quietly die: the 0.5% rate, the £3m threshold, the £15,000 allowance, and the rule that levy funds pay only for training and assessment. There are also two dates you should not conflate. The rename is April 2026. A separate set of funding-mechanic changes lands on 1 August 2026, the start of the 2026/27 funding year: the 10% government top-up on funds entering DAS is removed; fund expiry shortens from 24 to 12 months for contributions made from that date; and co-investment for a levy payer who has exhausted their pot rises from 5% to 25%, with no age scoping. Your existing pot and its expiry aren't affected retrospectively.

The "use it or lose it" clock is real, and shortening. But chasing it is exactly how you end up buying shelf-ware: training picked because the budget was expiring, not because an outcome was named. Here's the practitioner line on when urgency is allowed to drive timing. Urgency is a legitimate input when you already have a named cohort, a defined role, and a programme mapped to real work — then committing before the expiry is sensible treasury management. It is not legitimate when the cohort and the outcome are undefined; that is the scenario that produces shelf-ware. If you can't yet name what changed work looks like six months in, the right move is to let the funds expire rather than spend them on a programme built around a deadline.

Funding rules: what a data engineering programme can claim under the levy

A Level 5 data engineering apprenticeship is fully fundable from your levy pot as approved apprenticeship training against the Data Engineer standard (ST1386), plus its end-point assessment. If you're a levy-paying employer with funds in your account, that's what the funds are for.

A few rules shape what actually gets funded:

• Existing staff count. A member of your team can be the apprentice. Recognition of prior learning (RPL) then reduces the content, duration and price to reflect what they already know. RPL is a reduction mechanism, never an uplift, and it can't take a programme below the funding floor of 8 months, or 187 off-the-job training hours.
• Off-the-job training is per-standard. For new starts from 1 August 2025, the DfE publishes a minimum number of off-the-job training hours for each standard, replacing the old flat "20% of contracted hours" proxy. Build that protected learning time into the working week. It's not optional, and a reader who discovers it late walks away.
• The levy pays for training, not employment. Wages, travel and the like are yours.

If you've exhausted your levy pot, co-investment applies. Any start you plan now lands in the 2026-27 funding year (starts from 1 August 2026): non-levy employers are 100% funded for apprentices aged 16–24 and pay 5% for those 25 and over, while a levy payer with insufficient funds pays 25% — with no age scoping. "Non-levy" means an annual pay bill under £3m, never headcount (DWP apprenticeship funding rules, 2026-07-12).

For the full mechanics, the Growth & Skills Levy explainer and funding options go deeper than this section.

Who regulates it now: DWP funding policy and Skills England standards

Apprenticeship funding policy now sits with the Department for Work and Pensions (DWP). This was a machinery-of-government change — announced in the September 2025 reshuffle and formally confirmed by written ministerial statement on 16 September 2025 — not an agency being abolished. Higher education and under-19 skills stayed with the Department for Education (DfE). The 2026/27 funding rules are the first published under DWP ownership (GOV.UK apprenticeship funding rules, 2026-07-12).

Standards and assessment plans are Skills England's domain. Skills England replaced the Institute for Apprenticeships and Technical Education (IfATE) on 2 June 2025 — a separate event, under the DfE at the time; it then moved with skills policy to the DWP in September 2025. The old Education and Skills Funding Agency (ESFA) closed on 31 March 2025; if you see it referenced as current, that guidance is out of date. The levy itself is still collected by HMRC through PAYE, as it always has been.

The practical upshot: funding rules from the DWP, the standard your programme runs on from Skills England, collection through HMRC. None of this changes whether a Level 5 data engineering apprenticeship is fundable — it is — but it does change whose guidance you check when the details move.

What a Skills England–approved data engineering apprenticeship covers

The Data Engineer apprenticeship is a Level 5 standard, ST1386, approved for delivery and active — a real, Skills England–approved standard, not an in-development one (findatrainingprovider.co.uk, 2026-06-15). Level 5 sits at roughly foundation-degree level: above the Level 3 and Level 4 data qualifications, below a full bachelor's or master's.

What does a data engineer actually do?

A data engineer builds and maintains the systems that move data from where it's produced to where it's used — the pipelines that feed analysts, dashboards and AI models. When that layer is neglected, failures cascade downstream into broken reports and models nobody trusts. It's a distinct role from a data analyst: the analyst interprets and reports on data; the engineer builds the plumbing that gets clean, reliable data to them in the first place. Our what is data engineering guide and the data engineer vs data analyst comparison draw that line in full.

The current v1.0 of ST1386 applies to new starts up to 30 November 2026; a revised version applies to new starts from 1 December 2026. Learners who start on v1.0 before 30 November 2026 complete on v1.0 — the standard version does not change mid-programme. If you're planning a cohort that starts close to that date, confirm the start date with your provider before signing the commitment statement, and check the standard's current status and version on its Skills England page, since standards revise.

Our view on how this content lands: you get good at data engineering by building pipelines, not by studying them. On our Level 5 programme, learners work in Prism — a simulated e-commerce company built on 500M+ rows of real data — coached by people who've done the job, so the practice is on real data in a safe sandbox before it's on your production systems. The funding is plumbing; the method is the point.

UK data engineer pay: what the role commands

As a reported snapshot, an entry-level data engineer with under a year's experience earns average total compensation of around £31,758, and someone with 1–4 years earns around £42,853 (payscale.com, accessed 1 July 2026). Treat these as one reputable source's advertised averages, not a settled national figure. Pay varies widely by sector, region and stack, so check a current salary guide for your own market before quoting a number to a candidate or a board.

That steep rise with experience is what makes the role attractive to fund and to move into, and the reason it holds up is structural, not cyclical. Analytics and AI both depend on a working data layer, and the people who build it are scarce relative to the people who consume its output. That gap is why building data engineering capability in-house, rather than only hiring for it, tends to pay off.

How to set up a levy-funded data engineering programme with iO-Sphere

Start from the outcome, not the funding. Here's the order we'd run it:

1. Name the capability. What do you need these people to be able to do — build and maintain production pipelines, own a specific part of your data stack — and how will you see it in the work six months in? Write that down first.
2. Check your levy position. If your pay bill is over £3m, you're paying the levy and have funds in your DAS account. If you're a smaller employer, government co-funding covers most of the cost, and a larger levy-paying partner can transfer up to 50% of their unused funds to you (raised from 25% in April 2024). The employers page walks through both routes.
3. Judge the programme as if you were paying cash. Look at what learners will actually build, who coaches them, and how the practice maps to your stack. A levy-funded programme that doesn't change the work is still expensive; it spends working time.
4. Fit the funding and commission it. Confirm eligibility, agree the off-the-job training time, and start. Existing staff can be apprentices, and RPL trims the programme to what they don't yet know.

The reason to run it in that order is that the right route depends entirely on what you named in step one. If your data engineering need is building production pipelines from scratch, the Level 5 apprenticeship is the right route: 15 months of structured, coached practice on real data changes what someone can do, and a short course does not. If your team already has senior engineers and you only need tool upskilling, the new non-apprenticeship flexibility or an off-the-shelf short course beats a 15-month programme, because you'd be paying for time they don't need.

Where the levy-first approach quietly fails is the employer who picks the apprenticeship route because their DAS balance was expiring. They end up with a certificate and a learner who was never embedded in the right team, and the off-the-job hours become compliance theatre — protected learning time spent to satisfy a rule rather than to build a capability someone asked for. That's the failure mode naming the capability first is designed to prevent.

If you want to talk it through — your levy position, who's eligible, how the Level 5 data engineering apprenticeship maps to your team — that's the conversation to have next. Explore the programme, or talk to us about commissioning a cohort →

Common questions on levy eligibility and transfers

Can the Growth & Skills Levy fund a full data engineering programme?

Yes. A Level 5 data engineering apprenticeship, run against the Data Engineer standard (ST1386), is approved apprenticeship training and is fully fundable from a levy-paying employer's DAS account, together with its end-point assessment. The levy pays for the training and assessment, not for wages or other employment costs.

How long does a Level 5 data engineering apprenticeship take?

The ST1386 standard's typical duration is 24 months; iO-Sphere's delivery runs 15 months of training plus a 3-month end-point assessment. Recognition of prior learning (RPL) can reduce that where a learner already holds some of the required knowledge, skills and behaviours, but not below the 8-month funding floor. The current v1.0 of ST1386 applies to new starts up to 30 November 2026; learners who start on v1.0 complete on v1.0, and the version does not change mid-programme.

What are the off-the-job training requirements?

For new starts from 1 August 2025, the DfE publishes a minimum number of off-the-job training hours per standard, replacing the old "20% of contracted hours" proxy. This protected learning time must be agreed and planned into the working week before the programme starts. It's a condition of funding, not an optional extra.

Who pays the levy, and how much?

Employers with an annual pay bill over £3 million pay the Growth & Skills Levy at 0.5% of that pay bill, collected monthly through PAYE, with a £15,000 annual allowance offsetting the charge. These figures are unchanged from the old Apprenticeship Levy — the April 2026 reform renamed and redesigned the levy but left the core mechanic in place.

How much of our levy funds can we transfer to another employer?

Up to 50% of your unused levy funds can be transferred to other UK employers — typically supply-chain partners, SMEs or charities. That maximum was raised from 25% to 50% on 22 April 2024. This is how a larger levy payer can fund a data engineering apprenticeship at a smaller partner that doesn't pay the levy itself.

Can we put an existing employee on a funded data engineering apprenticeship?

Yes — existing staff can be apprentices. Where they already hold some of the required knowledge, skills and behaviours, recognition of prior learning (RPL) reduces the content, duration and price to match, though it can't take the programme below the 8-month / 187 off-the-job-hours funding floor. RPL only ever reduces; it never adds funding.

Who regulates apprenticeship funding and standards now?

Apprenticeship funding policy sits with the Department for Work and Pensions (DWP), following a machinery-of-government change formally confirmed on 16 September 2025; the 2026/27 funding rules are published under DWP ownership. Apprenticeship standards, including the Data Engineer standard, are the responsibility of Skills England, which replaced IfATE on 2 June 2025. HMRC still collects the levy through PAYE.

Can the levy fund shorter, non-apprenticeship training too?

From April 2026, phased in over time, up to 50% of the Growth & Skills Levy is intended to be usable on shorter, flexible non-apprenticeship training units rather than only full apprenticeships. The exact scope is still settling, so check the latest DfE/DWP funding rules before planning a non-apprenticeship spend. For a full data engineering programme, the apprenticeship route remains the straightforward one. iO-Sphere doesn't deliver apprenticeship units: the funding is small and the format constrains depth. A unit fits when an employer needs a fully-funded, short, narrow intervention and can't fund more — not where genuine capability-building is the goal.

Is iO-Sphere the right choice for every employer?

Not always. If you need Level 6, Level 7 or a data science degree apprenticeship, that's outside what we deliver — our Level 5 data engineering programme builds toward those but isn't a substitute, and you'd pursue the higher levels with a degree-apprenticeship provider. If you only need a one-week tool refresher for an already-capable team, a short course is a better fit than a 15-month apprenticeship. Where we're a strong fit is building genuine, applied data engineering capability in people from a wide range of backgrounds — see the data engineering hub for how the pieces connect.

Policy correct as of July 2026. Funding rules and effective dates change — confirm the current position with the latest DfE/DWP funding rules before committing.