Product leadership, before the job title
The AI Character Builder was not a one-off. These three projects show the same pattern in different school systems: a real problem, a prototype or process, a quality bar, and adoption by the people who had to live with it.
Whole-school timetable implementation
The problem. Balancing curriculum time, specialist lessons, teacher eligibility, rooming and whole-school constraints—including Arabic/Islamic staffing and avoiding poor subject sequencing.
My role. Led from architecture to deployment: SLT, principal, departments, HR/legal and the iSAMS handoff, with documentation and deadline management across three yearly cycles.
AI parent simulation for Head of Year recruitment
The problem. The school wanted to see how candidates would respond to a parent concerned about her child's academics. Role-play needed a staff member—and real parent interactions happen in writing anyway.
My role. Built the chatbot in a day, tested it, refined the parent behaviour and guardrails, won approval from the Head of Primary, and deployed it—with a scorecard out of five and detailed candidate feedback.
Culturally context-aware book suitability database
The problem. Teachers and librarians needed to know whether books were suitable for the local school context—culturally sensitive references such as pigs, alcohol, relationships or religious content.
My role. Built a database concept supporting search, suitability flags, notes and context-aware guidance—for teachers, librarians, and potentially schools across Qatar.
The key judgment: context-aware guidance beats a binary banned/allowed list—teachers need the why, not just the flag.
Different systems, same working loop—the one I'd bring to CoTeach.