The Birth of the App: From Sceptic to Advocate
TeachEdge.ai started as a practical response to the essay-marking workload — but quickly became a bigger idea: co-intelligence, where AI produces a strong first draft and teachers apply judgement, context, and care.
Quick Summary
- •TeachEdge.ai began as a solution to the sheer time cost of meaningful essay feedback.
- •The real challenge wasn't just the AI — it was building a teacher-and-student workflow that felt trustworthy.
- •Teacher oversight (reviewing and editing feedback) keeps the process human and improves quality.
- •The long-term direction is co-intelligence: AI and teachers augmenting each other rather than replacement.
Summary: I didn't build TeachEdge.ai because I wanted an "education revolution". I built it because I had too many essays, too little time, and a genuine curiosity about what AI could do. The surprise was how quickly it became a tool for co-intelligence — AI and teacher judgement working together.
Where the idea came from
The initial idea for the AI-powered essay app came from a place of necessity (and a dose of genuine interest in artificial intelligence).
As an economics teacher with hands-on experience marking A-level exams, I was all too familiar with the drudgery of essay marking. It wasn't just the monotony — it was the sheer effort required to provide meaningful feedback to every student in my class.
That's where my AI odyssey began: transforming my aversion to marking into a practical tool that could benefit both teachers and students.
The quest for something better than a shortcut
I didn't want a tool that simply made marking faster. I wanted something that genuinely extended what teachers are trying to do when we give feedback properly.
The objective was clear:
- build an AI application that serves as a reliable essay-marking assistant
- help students see AI as a collaborator, not a shortcut to academic success
The roadblocks and the revelations
This journey wasn't without its hurdles.
Building something teachers would actually use
Crafting an interface that catered to the intricate demands of both students and teachers required meticulous attention to detail.
It had to be:
- user-friendly (because no one has time for friction)
- capable of feedback with real depth (the "expert teacher with lots of time" problem)
- aligned to exam-board expectations, not generic comments
Changing student perceptions
One of the biggest challenges wasn't technical at all — it was student value-perception.
Students needed to see the app as an enhancement of teacher guidance, not a replacement. Without that, AI feedback risks being treated like a gimmick, or worse, a way to avoid thinking.
Keeping the teacher's touch
The hardest balancing act was maintaining the line between technological efficiency and the personal insight only a teacher has.
The app needed to increase productivity without disconnecting teachers from student progress.
That's why a key feature is the ability for teachers to:
- review the draft feedback
- edit it
- then return it to the student with confidence
What changed in the classroom
The app has drastically increased the number of essays students can write and receive feedback on.
But what stands out most is how it supports improvement — not just evaluation.
One feature I'm particularly proud of is the ability to rewrite sections of an essay in the student's voice, while also explaining the rationale behind each change. That's been especially helpful for students who would usually be working below a grade C, because it turns feedback into something they can immediately act on.
Contrary to my initial reservations, students were keen to engage with the AI feedback, and many became more willing to revise and resubmit.
A practical shift followed too: essays moved from homework towards classwork, partly to reduce overreliance on AI and keep the writing process firmly human.
The future is co-intelligent
It's worth acknowledging something slightly unnerving: AI's capabilities today are the most limited they will ever be.
The future isn't about building an autonomous educational ecosystem. It's about fostering co-intelligence — where AI and human educators augment each other.
That isn't a harbinger of teacher obsolescence. If anything, it's an evolution of the teacher role: less time spent on repetitive drafting, more time spent on judgement, coaching, and the parts of learning that are deeply human.
Final thought
This web application is more than a grading tool. It's a small proof of concept for what human–AI collaboration can look like in real classrooms.
It has changed how I approach teaching, and it has offered students a way to benefit from AI without losing the human element that defines good education.
As we keep refining this relationship, it's hard not to feel a bit excited about what other doors AI will open.
Gary Roebuck is Head of Economics at Holy Cross New Malden, and creator of TeachEdge.ai.
Related Posts
Why TeachEdge.ai Gets Marking Right (When Others Don't Quite)
Most AI marking tools are clever, but they often miss what real exam marking rewards: clear credit for what's there, even when answers are rushed. Here's how TeachEdge.ai is trained to mark like a teacher, not just apply a rubric.
Navigating AI Essay Marking and Feedback
AI can save hours on essay marking, but the real impact comes when a teacher reviews, tweaks, and stands behind the feedback. This post explains the 'human-in-the-loop' approach that makes AI feedback feel trustworthy and genuinely useful to students.
Stop Trying to Catch Them: Why AI Detection is a Dead End for UK Secondary Schools
AI detection tools cannot reliably prove whether a GCSE or A Level student used generative AI. Schools will get further by modelling good AI use and protecting supervised writing time.
Ready to transform your marking workflow?