Six evaluation areas. A year's worth of evidence. One document nobody has time to write.
Under the November 2025 inspection toolkit, Ofsted expects your self-evaluation form to reflect your service across six graded areas — leadership, inclusion, curriculum, achievement, behaviour, and routines — plus safeguarding. That's a living document, not something you update the week before they call. Most nursery managers write the SEF in a rush, drawing on memory rather than evidence, and the result reads like it. Your SEF agent reads your daily records, observation logs, incident reports, training matrix, and parent feedback — then drafts each section with evidence mapped to the toolkit's evaluation criteria. Your job is review and sign-off, not authorship.
What Your Agent Actually Does
Your SEF agent continuously assembles evidence from your existing records and drafts a self-evaluation that Ofsted actually wants to read — grounded in data, not aspiration.
Maps evidence to all six evaluation areas
Your agent reads observations, incident logs, training records, parent surveys, SEND plans, and safeguarding records — then maps each piece of evidence to the relevant evaluation area. Leadership, inclusion, curriculum, achievement, behaviour, routines — covered, with your actual data, not generic statements.
Drafts evidenced narratives, not wishful thinking
Each section gets a structured draft that references what you've done, not what you plan to do. Specific examples, real numbers, named improvements. The kind of self-evaluation that inspectors trust because it matches what they see on the floor.
Flags evidence gaps before Ofsted finds them
Thin on inclusion evidence? No recent parent feedback? Safeguarding training due for renewal? Your agent spots these gaps monthly and tells you in plain English — so you fix them when it's easy, not when it's urgent.
Tracks the new five-point grading scale
The toolkit grades each area from Urgent Improvement to Exceptional. Your agent assesses your evidence against the published criteria for each grade, so you know where you're sitting before the inspector arrives.
Updates itself as your records grow
The SEF shouldn't be a once-a-year exercise. Every new observation, incident, training completion, or parent comment is evidence. Your agent incorporates it continuously, so the document is always current — not six months stale.
The real numbers.
| Manager time writing and maintaining SEF | £1,500–£3,000/year |
| Consultant SEF review or support session | £500–£1,000 |
| Time gathering evidence from multiple systems | £500–£1,200/year |
| Realistic annual cost | £2,500–£5,200 |
| Agent build (one-off, configured to your records and systems) | £2,500–£4,000 |
| Monthly running costs (hosting + AI usage) | £100–£180/month |
| Toolkit updates as Ofsted evolves criteria | Included in first year |
| Realistic first-year total | £3,700–£6,160 |
The SEF is supposed to be a reflection of how well you know your own service. In practice, it's a document most managers dread — partly because it's a lot of writing, and partly because pulling evidence from five different systems is a chore nobody schedules time for.
The November 2025 toolkit makes this harder, not easier. Six evaluation areas with detailed criteria, a new five-point grading scale, and inclusion as a standalone area for the first time. Your agent doesn't make the framework simpler — it makes the writing and evidence assembly someone else's problem.
Good fit / not a fit.
This works brilliantly for:
- Any Ofsted-registered nursery or preschool preparing for the new inspection toolkit
- Settings where the manager writes the SEF and it hasn't been updated in months
- Multi-site operators who need consistent self-evaluation across settings
- Settings that have previously paid consultants to help with SEF preparation
This probably isn't for you if:
- Your setting is very small and the SEF is straightforward enough to maintain in an afternoon
- You have a dedicated quality or compliance lead with capacity for ongoing self-evaluation
- You're a childminder — the SEF requirements for group settings don't apply the same way
Handled like the sensitive data it is.
Children’s records — observations, safeguarding logs, SEND plans — are special category data under UK GDPR, and we treat them that way. Your agent runs on Claude via AWS Bedrock with an EU-only inference profile — meaning prompts and outputs never leave the EU, and are never used to train a model. Protected under Anthropic’s enterprise Business Associate Agreement.
Every run is logged for audit, every output is a draft your manager reviews and approves, and your data lives in a tenant isolated from every other customer. Parental consent verification is built into data workflows, and we’ll hand you a pre-filled DPIA template you can drop into your own records.
Need UK-only data residency? We offer an Azure UK South deployment as an enterprise add-on for customers with stricter procurement requirements. Full security details →
Let's talk.
We'll start by looking at your current SEF and the systems you use for daily records, observations, and incidents — what's exportable, what's paper-based, and where the evidence usually gets stuck. Usually a 15-minute conversation.
hello@nimblecroft.com