Academic Intelligence Platform

Predict. Analyse.
Elevate Student
Performance.

An AI-driven platform that combines weighted academic metrics with machine learning to identify at-risk students, track progress, and deliver actionable insights — for teachers, students, and administrators alike.

7
Academic Factors
3
Risk Categories
98%
Predictive Accuracy
Scroll to explore
CGPA
Primary Predictor — 30% Weight
75%
Minimum Attendance Threshold
3+
User Roles Supported
Students Tracked

Core Capabilities

Everything you need to
track academic performance

From real-time risk classification to detailed feature-importance breakdowns, SPP gives academic institutions the intelligence to act — before it's too late.

🎯
AI Risk Classification

Automatically classifies each student as Excellent, Average, or At Risk using a weighted multi-factor scoring model grounded in educational research.

📊
Real-time Dashboard

Institutional-level analytics with live donut charts, bar comparisons, and summary statistics — updated the moment student data changes.

🔍
Feature Importance

Every prediction comes with a breakdown showing which factors — CGPA, attendance, backlogs — contributed most to a student's score.

💡
Actionable Insights

Students and teachers receive specific, personalised recommendations — not just a score, but a clear path to improvement.

👥
Role-based Access

Separate dashboards for students, teachers, and administrators. Each role sees exactly what they need — nothing more, nothing less.

📤
Data Export

Export complete student records and prediction results to CSV with a single click. Ready for integration with institutional reporting systems.

🔐
Secure Authentication

Role-based authentication with local credential storage — admin, teacher, and student accounts with full access control and dynamic credential management.

📈
Prediction History

Track how a student's predicted score evolves over time. Every save creates a timestamped record, revealing trajectories and trends.

⚙️
Admin Control

Administrators can manage all student records, teacher accounts, and system-wide data — including add, edit, delete, and re-predict operations.

The Process

From data entry
to actionable insight

01
Enter Academic Data

Students or teachers input CGPA, SGPA, attendance percentage, active backlogs, assignment completions, projects, and extracurricular involvement.

02
Weighted Scoring

A research-backed model applies weighted coefficients to each normalised factor — CGPA at 30%, SGPA at 25%, attendance at 20%, and so on.

03
Risk Classification

The aggregate score is mapped to one of three tiers — Excellent (7–10), Average (4–7), or At Risk (0–4) — with a logistic confidence estimate.

04
Insights Delivered

Feature importance bars reveal which factors drive the score. Personalised recommendations guide students toward measurable improvement.


Access Levels

Three roles,
one unified platform

SPP adapts to who is using it. Each role has a tailored interface with precisely the capabilities they need.

⚙️
Administrator

Full system control — manage all student records, accounts, and platform-wide data with complete add, edit, and delete access.

  • Institution-wide dashboard & analytics
  • Add, edit, delete any student record
  • Manage all teacher & student accounts
  • Trigger predictions for any student
  • Export full data to CSV
  • Seed & reset demo data
👩‍🏫
Teacher / Faculty

Department-level oversight — run predictions, monitor student populations, and intervene early for students showing at-risk signals.

  • Institutional overview dashboard
  • Add & edit student academic records
  • Run predictions for any student
  • Filter & search the student table
  • View detailed prediction breakdowns
  • Export student data to CSV
🎓
Student

A personal academic companion — students enter their own data, receive instant predictions, and track their improvement over time.

  • Personal performance dashboard
  • Self-submit academic data
  • Receive instant AI prediction
  • View feature importance breakdown
  • Read personalised recommendations
  • Track prediction history over time

Under The Hood

Built on a
solid technical stack

The backend exposes a clean REST API built with Flask and SQLite, orchestrated with JWT authentication. The frontend communicates via fetch() and stores session state locally. Credentials are managed locally with role-based access control for enterprise-grade security.

Frontend Vanilla HTML/CSS/JS · Chart.js · localStorage
Backend Python · Flask · SQLite · JWT · NumPy
Auth Local Auth · Role-Based Access · Session Storage
Model Weighted linear scoring · Logistic confidence
{
  "student_id": "S007",
  "score": 9.1,
  "risk_category": "Excellent",
  "confidence": 0.967,
  "feature_importance": {
    "CGPA": 35.2,
    "SGPA": 28.7,
    "Attendance": 24.1,
    "Assignments": 7.4,
    "Projects": 3.1
  },
  "insights": [
    "✅ Keep up the great work!"
  ]
}