The SR Zone

NEB students in Nepal spend 12.7 hours/week preparing for board exams (source: 2024 SR Zone user survey, n=2,343) but lack visibility into their progress. Today's fragmented experience forces students to track studied topics in physical notebooks, attempt random practice questions without performance feedback, and rely on expensive last-minute tutors ($8-$15/hr) to identify knowledge gaps. This creates exam stress spikes: 68% report "feeling lost" about readiness two weeks before exams (source: same survey).

The AI Exam Dashboard addresses this with 120,000 active Class 11-12 SR Zone users (source: March 2024 MAU report) × 6X monthly practice quiz attempts (from current 1.2 to projected 7.2 attempts) × $5/month upsell potential ($3.60 saved tutor cost + $1.40 engagement premium) = $3.6M/year ARR upside. If adoption hits 40% of target: $1.44M/year. (Assumptions: 40% conversion from free to paid dashboard tier validated via concept testing with 127 students).

This feature IS a closed-loop exam prep system with automated gap detection. It IS NOT a general study planner, peer comparison tool, or AI tutor replacement — it focuses exclusively on converting existing SR Zone content into measurable progress signals.

Our wedge is hyper-localized NEB topic alignment because our practice sets are built by Nepali examiners while competitors use generic CBSE/NCERT content.

Quantified Baseline

Business case: 120,000 students × 60% tutor cost avoidance × $100 avg panic spend = $7.2M/year recoverable value.

Behavioral Root Cause Students currently treat SR Zone as a PDF library rather than a prep system because there's no feedback loop between studying and exam readiness. The JTBD: "When I'm preparing for board exams, I want to know exactly which chapters need more work so I can study efficiently instead of guessing."

##SECTION:solution_design``` ┌───────────────────────────────────────────────┐ │ Exam Progress Dashboard 🕒12d left │ ├───────────────────────────────────────────────┤ │ Physics ████████░░ 72% Weakest: Optics │ │ Chemistry █████████░ 85% Weakest: Thermodynamics │ │ Math ████░░░░░░ 40% Start ASAP → │ └───────────────────────────────────────────────┘

┌───────────────────────────────────────────────┐ │ PANIC MODE: Weakest 3 Topics [START] │ ├───────────────────────────────────────────────┤ │ 1. Optics (Physics) - 5 key derivations │ │ 2. Probability (Math) - 3 practice sets │ │ 3. Thermodynamics - 2 previous exam papers │ └───────────────────────────────────────────────┘

**Design Decisions**
1. Decision: Progress tracking scope
Chose: Auto-track via existing practice sets
Rejected: Manual entry or new content
Rationale: Reduces friction; leverage existing 8,200 practice questions

2. Decision: Weakness detection method
Chose: ML model on practice set errors
Rejected: Simple % correct thresholding
Rationale: Identifies concept-level vs question-level gaps

Phase 1 — MVP (10 weeks) US#1 — Progress Visualization

• Given a student completes Physics Ch1 practice
• When they view dashboard
• Then completion % updates within 2s (P1: 99.5% accuracy)

US#2 — Panic Mode Activation

• Given <24h before exam
• When student clicks Panic Mode
• Then checklist includes top 3 weak areas from last 7 days (P0: 100% accuracy)

Out of Scope

What We Are NOT Measuring:

• Total users (measures vanity, not engagement)
• Social shares (non-core to JTBD)

Risk: Students distrust AI recommendations Probability: Medium Impact: High Mitigation: Show "Why This Topic?" explainers with textbook page references (Owner: UI Lead B. Thapa by 8/15)

Risk: Nepal Data Protection Act compliance Probability: Low Impact: Critical Mitigation: Anonymize all practice data before model training (Owner: Legal R. Shrestha by 7/30)

Kill Criteria:

1. <2,000 Panic Mode activations by first exam cycle (Oct 15)