Built from a real classroom

The story

MyTeachingPal began with a single Year 12 Chemistry teacher spending an entire weekend marking 27 draft research investigations. Each student submission required the same process: read the rationale, check the combustion equations, verify the research question was open-ended, inspect all three sources for peer-reviewed citations, evaluate each limitation correctly, check the conclusion restated the RQ with specific values — and write 200 carefully calibrated words of feedback in the margins.

After building and refining a workflow over many hours across many conversations with AI tools, the right approach became clear: the AI needed to know the specific chemistry errors to flag (CH₄ is not a diesel equation), understand the exact comment style (“not a limitation” not “this doesn't quite qualify as a limitation because...”), and produce output that looked identical to teacher-written Word comments — not a PDF, not an HTML file, but actual Word Review comments in the student's own document.

MyTeachingPal is that workflow, productised. It handles the first pass — which covers most of the marking work — so the teacher only needs to review, adjust, and approve. The result: what used to take a weekend now takes a morning.

It grew from there. MyTeachingPal now does full QCAA final marking too — a three-step best-fit against the ISMG, with a highlighted ISMG and an annotated student response you can hand straight back. And because no two markers are identical, you can train MyTeachingPal on your own marking: it compares its judgement to yours and folds your reasoning into every future mark.

We're expanding to more subjects, year levels, and assessment types — but the commitment stays the same: marking that sounds like the teacher and defers to their judgement, not a chatbot's.