AI in Education
Turn AI possibilities into everyday academic practice: faculty capability, dependable workflows, and honest evaluation of what actually helps.
What AI should change, and what it must not
AI should take routine work off faculty desks: first drafts of question banks, formatting, checking, summarising. That frees the hours that matter: explaining an idea well, noticing the student who is stuck, deciding what to teach next.
It must not replace judgement. Every AI-generated solution is checked by a person who can do the mathematics; nothing reaches a student unverified; and a workflow earns its place only if it saves time without costing accuracy.
What I’ve done
Train
I train 100+ faculty and subject-matter experts in prompting, verification and LaTeX-first content production.
Deploy
I put AI-assisted pipelines into question banks, solution books, slide decks and question-paper verification, with custom agent skills built to each department’s standards.
Optimise
I use AI on routine academic work (content, verification, analysis) so faculty time goes into teaching and outcomes.
How I show evidence
- 1. DemonstrationPublic demos, clearly labelled as demos.
- 2. Workflow accountHow it runs in real academic work, with its limitations.
- 3. Measured resultOnly with a documented baseline to compare against.
- 4. External feedbackFrom faculty and institutions, shared with their permission.
No rung is skipped: a demo is never presented as a result.