MVP and AI System Clarification
Chapter 11 of 15

CHAPTER 11: MINIMUM VIABLE PRODUCT (MVP) DOCUMENTATION AND AI SYSTEM CLARIFICATION

11.1 Chapter Introduction

This chapter defines the Minimum Viable Product (MVP) for the Intelligent Learning Management System (ILMS). The MVP represents the first production-intended release, built upon insights gained from the prototype phase.

In addition, this chapter explicitly clarifies the role of Artificial Intelligence (AI) within the system, addressing potential ambiguity and ensuring the AI design is accurately understood, scoped, and defensible.


11.2 Purpose of the MVP

The MVP exists to:

  • Deliver real institutional value
  • Operate reliably at scale
  • Validate adoption by students and lecturers
  • Provide measurable academic outcomes
  • The MVP is not experimental; it is operationally stable but intentionally limited in scope.


    11.3 MVP Scope Definition

    11.3.1 Core Features Included

    The MVP includes:

  • Secure authentication and identity resolution
  • Role-based dashboards (Student, Lecturer, Admin)
  • Automated academic placement
  • Course and unit management
  • Attendance tracking (online and physical)
  • Assessment submission and grading
  • Skill DNA analytics (advisory)
  • Reporting and audit logs
  • 11.3.2 Features Deferred Beyond MVP

    Deferred features include:

  • Advanced predictive analytics
  • External employer integrations
  • Cross-institution data sharing
  • Financial and billing modules

  • 11.4 System Architecture at MVP Stage

    The MVP architecture maintains:

  • Modular services
  • Central identity resolution
  • Separated analytics layer
  • Prototype shortcuts are removed, and data persistence is enforced.


    11.5 AI System Clarification

    11.5.1 Is the ILMS an AI System?

    Yes — the ILMS contains AI-assisted components, but it is not a fully autonomous AI system.

    The system uses controlled, narrow AI techniques to support analysis and decision assistance, not seen as replacing human judgment.

    11.5.2 Where AI Is Used

    AI is applied in the following areas:

  • Skill DNA inference
  • Pattern recognition in engagement data
  • Trend analysis across assessments
  • These components operate on historical and real-time academic data.

    11.5.3 Where AI Is NOT Used

    AI does not:

  • Assign grades
  • Make pass/fail decisions
  • Replace lecturers
  • Enforce disciplinary action
  • Human authority remains central.


    11.6 AI Design Philosophy

    The AI system follows these principles:

  • Explainability over complexity
  • Assistive intelligence, not automation
  • Transparency and auditability
  • Ethical constraint enforcement
  • Black-box decision-making is explicitly avoided.


    11.7 Skill DNA as an AI-Assisted Module

    Skill DNA uses:

  • Rule-based inference
  • Weighted indicators
  • Confidence accumulation over time
  • This qualifies as augmented intelligence, not general AI.


    11.8 Data Requirements for MVP AI Components

    Required data includes:

  • Attendance records
  • Assessment outcomes
  • Engagement metrics
  • No biometric or sensitive personal data is used.


    11.9 Performance Expectations

    The MVP is expected to support:

  • 1,000–15,000 concurrent users
  • Semester-scale data volumes
  • Near real-time dashboard updates

  • 11.10 Governance and Control

    AI outputs are:

  • Viewable
  • Interpretable
  • Non-binding
  • Institutional policies override all analytics.


    11.11 MVP Validation Criteria

    Success metrics include:

  • System uptime
  • User adoption rates
  • Accuracy of academic mapping
  • Lecturer and student feedback

  • 11.12 Transition Beyond MVP

    Post-MVP development may introduce:

  • Enhanced analytics
  • Personalization engines
  • Controlled predictive modeling
  • These are conditional on governance approval.


    11.13 Chapter Summary

    This chapter defined the MVP scope, clarified production readiness, and explicitly explained the AI role within the ILMS. The system is AI-assisted, ethically constrained, and human-governed, making it suitable for real-world academic deployment.