v2.0 Unified Architecture

Adaptive Challenge Engine

QuestSim Platform — Pythagoras Quest — Three-Voice Narrative Architecture

🎮 Want to see the ACE in action? Try the interactive prototype — play as a student or switch to teacher view.
▶ Launch Pythagoras Quest Prototype
Architecture
Three-Voice Model
Scaffolded Authoring
UX Screens
Roadmap

The System

Detects what happened

Pattern detection, mastery signals, engagement velocity, error classification, pathway transitions

📋

The Teacher

Defines what it means

Pedagogical context, narrative framing, tone selection, assessment criteria, feedback messages

💭

The Student

Shares how it feels

Math confidence, self-assessed understanding, engagement readiness, session mood

▼   THREE-VOICE SYNTHESIS   ▼
👤

Learner Profile

+ Affective Layer
Knowledge state, velocity, engagement + self-reported emotions
🎯

Challenge

Bloom's taxonomy levels, difficulty vectors, extension variants
🗺️

Mastery Map

+ Calibration Gap
Concept graph, confidence scores, decay modeling
🛤️

Difficulty Pathway

+ Narrative Layer
Foundation → Extension → Mastery with teacher narratives
🔄

Feedback Loop

+ 3-Voice Synthesis
Event-driven engine with narrative response composition

Three-Voice Narrative Flow

How the same performance event produces different narratives based on three information sources.

📊

Performance Event

Student answers a challenge (correct or incorrect)

System Analysis

ACE evaluates accuracy, time, error type, mastery state

💭

Affective Query

Check student's self-reported confidence and mood

📋

Narrative Match

Find teacher's rule matching this combination

🔍

Calibration Check

Cross-reference self-assessment with Mastery Map

💬

Narrative Delivery

Synthesized feedback: data + meaning + emotion

Same Event, Four Different Stories

Student gets a Pythagorean theorem problem wrong →

😰 Anxious Student

Confidence: Storm cloud | Understanding: Confused | Readiness: Calm

"Math can feel tough sometimes, and that's completely normal. Let's look at this together — try labeling the sides first."

🏃 Rushing Student

Confidence: High | Time: 12s (median 45s) | Readiness: Maximum

"You dove in fast — I like the confidence! But this one has a twist. Take another look at which side is the hypotenuse."

🎭 Underconfident High Performer

Confidence: Storm cloud | BUT Mastery Map: 0.85 | Recent: 4/5 correct

"You said math feels hard right now, but your recent work tells a different story. You got 4 of 5 right. Trust the evidence."

🤔 Genuinely Confused

Confidence: Neutral | Understanding: Confused | Mastery Map confirms gap

"This builds on right triangle identification. Let's go back for a few minutes — once that clicks, this will make more sense."

Scaffolded Authoring System

Solving the tacit knowledge problem: teachers author narratives at their comfort level.

"Even with a tool like this, teachers still have to put their pedagogical thinking into words clearly enough to build the rules. A lot of what experienced teachers do is pretty intuitive." — The tacit knowledge insight

Three Authoring Approaches

01

Scenario-First

"What would you say to this student?"

  • System shows concrete student data
  • Teacher responds naturally
  • System structures it into a rule
  • Activates tacit knowledge directly
02

Teaching-Moment Templates

Organized by how teachers think

  • "Student who's struggling"
  • "Student who rushed"
  • "Student who seems anxious"
  • Rule logic hidden by default
03

Watch-Me-Teach

System learns from teacher's practice

  • Teacher gives feedback naturally
  • System identifies implicit rules
  • Teacher reviews and approves
  • Intuition → automation

Progressive Disclosure Levels

L1

Template User

Select templates, customize messages, preview. No rule logic visible.

L2

Template Customizer

Modify triggers, combine templates, see readable rule logic.

L3

Rule Author

Create from scratch, complex multi-condition triggers, full variable library.

UX Screens Mapped to Objects

Every screen corresponds to a primary OOUX object. Gold border = new in v2.0.

Student Dashboard

Primary: Learner Profile

Launch simulation, view progress, review personal data stories, see mastery trajectory and affective trends.

Pre-Game Check-In NEW v2.0

Primary: Learner Profile (Affective)

Pythagoras NPC asks 3 questions: math confidence (emoji scale), self-assessed understanding (visual metaphors), readiness (adventure paths). 60-90 seconds.

Simulation View

Primary: Challenge

Interactive challenge with scaffolding, real-time three-voice narrative feedback, mid-session pulse checks at transition points.

Progress Map

Primary: Mastery Map

Concept graph with mastery levels, teacher-authored contextual labels, calibration gap indicators for student self-awareness.

Pathway Designer

Primary: Difficulty Pathway

Author tiers, set transition thresholds, assign challenges. Three tabs: Challenges, Thresholds, Narrative.

Narrative Editor NEW v2.0

Primary: Feedback Loop (Narrative Config)

Narrative rule cards, template library by teaching moment, scenario-first builder, live preview with sample data, progressive disclosure levels.

Session Reflection NEW v2.0

Primary: Feedback Loop (Student Output)

End-of-session data story combining mastery progress, performance narrative, teacher-authored reflective prompts, forward-looking message.

Analytics / Debrief View

Primary: Feedback Loop (Aggregated)

Cohort profiles, pathway performance, aggregated affective trends, class-level data stories, debrief mode for classroom discussions.

Unified Implementation Roadmap

Four phases delivering Core ACE, TNE, and SSNE capabilities together.

Phase 1: Foundation

Q2 2026
Core ACE

Learner Profile, Challenge, basic Mastery Map, simplified Feedback Loop

TNE

Template Library. Select and customize pre-built narrative templates.

SSNE

Pre-game check-in with default NPC dialogue. Three dimensions.

Phase 2: Adaptive Pathways

Q3 2026
Core ACE

Difficulty Pathway, tiered challenges, velocity tracking, tier transitions

TNE

Custom Rule Builder + Live Preview. Scenario-first authoring. Levels 1–2.

SSNE

Affective-aware rules. Calibration gap detection. Mid-session pulses.

Phase 3: Intelligence

Q4 2026
Core ACE

Full difficulty vector, mastery decay, preference inference, analytics

TNE

Class dashboard. Debrief mode. Watch-Me-Teach. Narrative sharing.

SSNE

Aggregated affective trends. Pastoral view. Wellbeing notifications.

Phase 4: AI-Assisted Refinement

Q1 2027
Core ACE

Self-improving pathways. Cross-cohort learning. Gap detection.

TNE

AI narrative suggestions. Teacher retains editorial control. Level 3.

SSNE

AI affective-pattern suggestions. Longitudinal trend analysis.