PRAGYAAi☀️
Foundations for a Brighter Tomorrow
AI-powered, evidence-driven FLN diagnostics that help teachers understand, support, and strengthen every child's learning journey in Mathematics.
The Classic Regrouping Error: 83 − 47 = 46
Student borrowed 10 into ones ($13 − 7 = 6$), but forgot to decrement the tens digit ($8 − 4 = 4$ instead of $7 − 4 = 3$). PRAGYA detects the slip mathematically, isolates 2 verdict-bearing checks, enables teacher verification, and coordinates remedial activity Borrow & Build.
The 7 Harmonized Steps of PRAGYA
Every step connects with verified contracts and persists state across the entire classroom journey.
Assessment
🟦 Ankit (Input)Fixed 5-item subtraction assessment with manual or simulated OCR input.
OCR & Trace
🟨 Abhay (Intelligence)Decomposes student responses into mathematical error signatures.
Diagnosis
🟨 Abhay (Intelligence)Two verdict-bearing checks + one warm-up baseline check isolate the gap.
Teacher Verify
🟪 Abhinav (Action)Teacher exercises ultimate agency: [ Accept ] [ Reject ] [ Change ].
Group
🟪 Abhinav (Action)Students grouped deterministically around confirmed learning gaps.
Intervene
🟪 Abhinav (Action)Concrete-Representational-Abstract (CPA) manipulative activities.
Reassess
🟪 Abhinav (Action)Same-skill measurement with raw before/after counts (+60% net growth).
Measurable Progress
Before vs After comparison calculated strictly from counts: 2/5 (40%) → 5/5 (100%) = +60% FLN improvement.
Deterministic Rule Engine
No generative LLM guessing scores, verdicts, or evidence. All diagnostic checks are mathematically computed in pure functions.
Teacher in the Loop
PRAGYA provides evidence and suggestions; the teacher has ultimate authority to Accept, Reject, or Change the confirmed gap.
CPA Remedial Pedagogy
Targets the root misunderstanding using physical base-10 manipulatives (Concrete), place value charts (Representational), and standard subtraction (Abstract).