When we reflect on our own teaching, context is everything.
The lesson you teach to a top set first thing on Monday is not the lesson you teach to a mixed-attainment Year 9 class last period on Friday. Different prior knowledge. Different pacing. Different personalities, routines and pressure points. Every teacher knows this instinctively. Yet most coaching, and most observation, has struggled to hold that nuance.
That is the problem we have just set out to solve. We have shipped Teaching Classes, a way for Starlight to understand the specific groups you teach, so the feedback you receive becomes more accurate, more relevant and more useful over time.
Teaching is contextual. Coaching should be too.
One of the quieter problems in professional development is comparison. When you receive feedback on a lesson, what exactly is it being measured against?
Until now, Starlight compared each lesson against all of your previous lessons. That is useful for spotting your habits as a teacher, but it meant a demanding Year 8 group could end up sitting alongside a confident Year 11 set, or a one-off cover lesson blended into a carefully sequenced unit. It is the same trap we have written about with teacher talk time, where a single tracked number flattens a picture that was never that simple. That is not how teaching works, and it is not a fair basis for reflection.
With Teaching Classes, Starlight now keeps those comparisons where they belong. You can group your recordings into named classes, such as 7TE, 9 History or 11A Science.
From that point on, Starlight tracks lesson progression, topic coverage, difficulty and pupil development within that class only. Feedback starts to follow the genuine arc of a teaching journey rather than a blend of unrelated groups.
What changes, and what stays the same
There is an important distinction worth being clear about. Your teaching technique is still drawn from across everything you teach. Questioning, explanation, modelling, routines, the way you handle behaviour: these are part of who you are as a teacher, and your development in them shows up across all your classes.
What changes is the contextual comparison. When Starlight reflects on progression, prior learning, sequencing and where a group is heading, it now holds that inside the relevant class. Put simply, Starlight still learns you across all your teaching, but it learns your classes separately. That sounds like a small distinction. In practice it is the difference between feedback that feels generic and feedback that feels like it was written about the lesson you actually taught. Which is, in the end, what teachers tell us they value most.
For most teachers, it simply happens
Teachers do not need another system to manage, so for most people there is nothing to set up. When you upload a lesson or cover lesson, Starlight groups your recordings by year group and subject automatically and creates the class for you in the background. Teach one Year 7 Science class and those lessons are already grouped together, with the feedback benefiting straight away. Your existing recordings are included too, so nothing from your history is lost. Past lessons that match on year group and subject are folded into the right class.
The one moment you might want to step in is when you teach more than one class in the same year and subject, say both 7TE and 7KI Science. Starlight cannot reliably tell those apart from the audio alone, and we made a deliberate choice not to guess. Rather than risk inaccurate feedback, we leave those lessons unassigned until you confirm the class yourself. Accuracy, and the trust that depends on it, matters more than a confident-looking assumption. A quick manual assignment is all it takes, and from then on your choice always wins. Automatic matching will never override it.
If it helps, create a separate class for each group you teach, and your feedback stays cleanly comparable.
A clean slate every September
Teaching resets every year. New classes, new dynamics, new names on the register. So Teaching Classes reset automatically each academic year. Your Year 8 class this year is not your Year 8 class next year, even when the timetable looks familiar. From September, every class starts fresh. Last year’s cohort never bleeds into this year’s feedback, and your class list stays focused on the groups in front of you right now. You do not need to do anything for that to happen.
A small feature solving a bigger problem
At first glance this might read like a minor update. It is not. It speaks to one of the hardest questions in AI coaching: how do you make feedback feel genuinely personal and context-aware, at scale?
Coaching has always been powerful, and always been hard to scale. A formal observation tends to happen once or twice a year. Starlight works on a different principle, that feedback matters most when it is specific, timely, actionable and regular.
But frequency only counts for something if each piece of feedback stays meaningful.
Teaching Classes is a step towards that, a coaching layer that understands not teaching in the abstract, but your teaching, with your groups, tracked honestly over time.
Because real growth was never about a single lesson. It is about patterns, about progress, and about the small improvements that compound across weeks, months and years. That is what Starlight was built for.
If you would like to see how this works in practice, you can book a short demo.
Spark Insight with Starlight, and bring sharper context to the teaching journeys that matter most.
🎥 Subscribe to our channel here: https://www.youtube.com/@Star21-ai
🌐 Read more on our blog: starlightmentor.com/blog
💡 Explore the platform: www.starlightmentor.com
🐦 Follow us on X: @star21starlight
The Insight Engine is written by Adam Sturdee, co-founder of Starlight, the UK’s first AI-powered coaching platform, and a senior leader with responsibility for teaching, learning and coaching. This blog is part of a wider mission to support educators through meaningful reflection, not performance metrics. It documents the journey of building Starlight from the ground up, and explores how AI, when shaped with care, can reduce workload, surface insight, and help teachers think more deeply about their practice. Rooted in the belief that growth should be private, professional, and purposeful, The Insight Engine offers ideas and stories that put insight, not judgment, at the centre of development.
