As the academic year draws to a close, I find myself doing something I rarely make time for: looking back. It’s been, without a doubt, one of the most rewarding and intellectually rigorous years of my career.
In the world of special educational needs and disabilities (SEND) education, the temptation to jump on the latest tech trend is constant. But my philosophy has always been firmly pragmatic: if it doesn’t solve a concrete problem, it’s just noise. This year, my focus remained steadfast on that principle, whether I was navigating the mechanics of a student-led business or engineering new workflows for classroom evidence.
Beyond the Classroom: The Junction Printshop
A major highlight has been watching the Junction Printshop evolve into a thriving student-led enterprise. We started with very little formal infrastructure, relying instead on a cocktail of experimentation, grit, and genuine teamwork.
There is a distinct scientific beauty in watching a group of learners dismantle a problem—whether it’s a print run that isn't aligning or a deadline that’s looming—and finding their own solution. Seeing these students cultivate confidence, independence, and the kind of employability skills that don't appear on a standard curriculum has been incredibly fulfilling. The true measure of our success? Seeing several of these learners secure supported internship placements for next year in an increasingly tough job market. It’s a testament to the fact that when we provide meaningful, real-world experiences, our learners don't just participate; they thrive.
Engineering Accessibility: The Birth of CSST-AI
My background in scientific research often informs how I approach teaching. When I see an inefficiency, I don’t just want to patch it; I want to engineer a better system. This year, that drive led to the development of Contextualised Speech-to-Structured Text AI (CSST-AI).
Like many educators, I’ve long struggled with the sheer volume of administrative documentation required to track learner progress. CSST-AI began as a selfish workload solution: a workflow combining wireless microphones, speech-to-text transcription, and carefully engineered prompts to turn messy, spoken classroom observations into clean, structured learner evidence.
Yet, as is often the case with good tools, it evolved. It transformed from a teacher’s timesaver into an accessibility powerhouse. It’s now helping learners find their own voice, assisting in the creation of dynamic CVs, and opening doors for job applications. It’s a reminder that good tech doesn’t replace human connection; it provides the scaffolding necessary for that connection to flourish.
Sharing the Practice
It has been a privilege to share these developments nationally. Presenting at the Natspec Peer Exchange and delivering workshops on Assistive Technology has been a highlight, but I’ve also been heavily involved in the broader conversation around ethical AI policy within educational trusts.
Perhaps most rewarding, however, was having this work featured in Natspec Quality Times. There is something uniquely humbling about seeing an idea born in the grit of day-to-day classroom practice recognised on a national stage. It reinforces the idea that we, as educators, are the best researchers of our own practice.
Looking Ahead
Innovation is a hollow word if it isn't anchored in impact. Whether it’s the printshop floor or the digital architecture of our classroom evidence, the common thread this year has been removing barriers rather than creating them.
I am deeply grateful to the colleagues, learners, and organisations who have supported and challenged me these past few months. As I look toward the next term, I am more convinced than ever that innovation matters most when it creates better opportunities for our learners to live independent, successful lives. That, at the end of the day, is the only metric that truly counts.
"This text was conceived and directed by a human, using Voice-to-Text and AI assistance to overcome a dyslexia induced literacy barrier."
