Why teachers need explainable AI, not just accurate AI — building the KC dashboard
NumPath's teacher dashboard previously showed one number per student: 7-day accuracy. A teacher looking at "Emma — 43%" had no idea whether Emma was struggling with borrowing, place value, number sense, or all three. The number was technically correct and completely unactionable. The team added a Knowledge Component (KC) mastery panel with color-coded progress bars per skill that expand to show p_mastery %, mastery level label, and opportunity count. The backend is a single endpoint backed by a left-join. The team chose option 3 of three options: accuracy-only (fast but unactionable), raw BKT parameters (complete but overwhelming), or KC mastery levels (translate p_mastery into a three-tier label: Novice / Developing / Mastered). The mastery level thresholds are defined as named constants in get_kc_states.py.
Explainable AI turns opaque accuracy scores into actionable insights for educators.