Five reports in 5 hours: what AI-augmented inspection prep actually looks like

Adam Sturdee
Mar 29
5 min read

Data doesn't gather itself. The schools that will thrive in the next decade of accountability are the ones quietly building their digital core right now.

When a school has built its digital core, an inspection is no longer a sprint to gather evidence. It becomes a structured conversation with data you already own — and AI can help you shape that conversation with precision and speed.

Last week, a school I work with faced an imminent inspection. Not Ofsted — a Section 48 Catholic inspection, which interrogates Catholic Life and Mission, Religious Education, and Collective Worship with a particular granularity that only inspectors who understand that world can fully appreciate.

In the day before the team arrived, we produced five substantive reports. Not drafts. Not slide decks with bullet points. Properly written, evidence-led documents that a senior leader, a governor, or an inspector could pick up and navigate without a guide.

5 formal reports produced

844 student survey responses synthesised

214 anonymised coaching log entries analysed

Those five reports drew on survey data from 844 students across seven year groups, 214 entries from a professional coaching log (all anonymised), a year's worth of assembly observation records, programme documentation, and live AI coaching data. Each one had a distinct audience, a distinct narrative purpose, and a distinct evidential structure.

I want to be honest about what made this possible — and what it reveals about the emerging relationship between school leadership, data infrastructure, and AI.

The digital core is the prerequisite

None of this works without the underlying data. Before AI can help you write a report, someone has to have built the habit of recording things — coaching conversations, student voice responses, assembly observations, programme outcomes — in a consistent, retrievable format.

What AI changes is the speed and quality of synthesis. Where a senior leader might spend three days pulling together a coherent narrative from a year's worth of fragmented records, AI can help structure and draft that narrative in a fraction of the time — leaving the human to focus on accuracy, tone, and nuance.

The reports produced weren't AI-generated in any crude sense. They were AI-assisted: the data, the professional judgements, and the institutional knowledge were human. The synthesis, the structure, the language — that's where the collaboration happened.

What the reports actually covered

Each document was tailored to a specific purpose and a specific reader. That specificity is what makes them defensible as evidence rather than just window dressing.

Coaching programme report for RE — drawing on structured coaching data to demonstrate the quality and consistency of professional development in the RE department, framed carefully to distinguish coaching from performance management.
Assembly observation report — synthesising entries from the headteacher's coaching log to evidence the depth and character of collective worship across the school year.
P4C student voice report — mapping 844 survey responses across three terms to the principles of Catholic Social Teaching, showing how philosophical enquiry was developing students' ethical reasoning and spiritual literacy.
Words, Wisdom & Worship programme evidence report — documenting how a bespoke whole-school literacy and spirituality programme integrated shared reading, prayer development, and Catholic Social Teaching.
AI coaching data report — produced using a refined prompt developed collaboratively with a senior colleague, synthesising live data from the school's AI-powered coaching tool to present a picture of teaching and learning quality.

Each report existed to do one thing: give the inspection team a clear, evidenced, well-structured account of something the school was genuinely doing well. Not spin. Not aspiration. Evidence.

"Inspection readiness isn't about performing for two days. It's about having a coherent narrative that accurately represents the school you've built — and being able to produce it on demand."

The sensitivity no one talks about

One of the recurring challenges in this work was framing. Specifically: how do you present coaching data without it reading as quality assurance data?

This matters enormously in professional cultures where trust is fragile. Coaching logs exist to support teacher development — not to generate accountability records. The moment a coaching document starts reading like an appraisal file, the professional relationship it's supposed to support begins to erode.

Getting the framing right required careful prompt engineering and several revision cycles. The final documents were explicit about purpose, careful with interpretive language, and structured to foreground professional learning rather than compliance. That kind of nuance is where the human-AI collaboration earns its keep. And the evidence base, analysis and final reports were always anonymised and aggregated.

Speed as a genuine differentiator

I want to be direct about something the educational world sometimes finds uncomfortable: speed matters.

An inspection call with short notice doesn't give you three weeks to commission a consultant and iterate through eight drafts. It gives you days — sometimes hours. The schools that will navigate the coming decade of inspection well are those that can mobilise their evidence base rapidly, shape it into a coherent narrative, and put it in front of the right people without losing either accuracy or quality.

That's exactly what AI-augmented report production enables — once the digital infrastructure exists to make it possible. The investment isn't in the AI tools themselves. It's in the years of consistent data practice that give those tools something real to work with.

What this looks like for other inspection frameworks

The approach I'm describing isn't specific to Catholic inspections. The same logic applies to Ofsted's new report card framework, to ISI inspections in the independent sector, to any accountability process that asks a school to demonstrate the quality of its practice through structured evidence.

The variables change. The narrative architecture changes. But the underlying method — drawing on a school's digital core, structuring the evidence, crafting a targeted report for a specific audience — stays consistent.

The question for every school leader, regardless of sector, is the same: if an inspection team arrived tomorrow, how long would it take you to put a coherent, evidenced account of your school's strongest work in front of them?

If the answer makes you uneasy, the conversation about digital infrastructure and AI-augmented leadership probably needs to happen sooner than you think.

Interested in this kind of work?

I work with school leaders on digital strategy, inspection readiness, and AI-augmented leadership practice. If you're thinking about how to build your school's evidential infrastructure — or how to use it more effectively — I'd be glad to talk.

Adam Sturdee is a senior leader and co-founder of Starlight, the UK’s teacher-first AI-powered transcript-based coaching platform for educators. His work sits at the intersection of dialogic practice, instructional leadership and responsible AI strategy for schools and trusts.

He will be presenting his research on AI-supported coaching at the BERA TEAN Conference 2026: https://www.bera.ac.uk/conference/bera-tean-conference-2026