Research-Oriented Semester Project Program

16-Week Hybrid Agile-Research Program for B.Tech CSE Students

Gandhi Engineering College (Autonomous)
After College Hours

16 Weeks

6 Structured Phases

Max 4 Members

Team-Based Projects

60 Projects

Across 6 Domains

3 Tracks

Ability-Based

Program Structure

Hybrid Agile-Research Methodology

Every project is decomposed into two parallel components: Software Component (implementation) and Research Component (evaluation & analysis). Both converge on a single results-and-documentation deliverable.

📋 Phase 1-2

Problem Definition & Research

Weeks 1-5

⚙️ Phase 3-4

MVP & Iterative Development

Weeks 6-10

🎯 Phase 5-6

Evaluation & Finalization

Weeks 11-16

📊 The Evaluation Discipline

Phase 5 (Weeks 11-13) requires every team to answer 7 questions:

1. What did you build and what problem does it solve?

2. What is your baseline (existing solution/simple model)?

3. How does your solution compare to the baseline?

4. How did you test/evaluate your system?

5. What metrics did you use and why?

6. What were your main error categories?

7. What would you do differently with more time?

Student Tracks

Track A
~50% of students

Application Development

MERN / Python Full-Stack

• College ERP Systems

• Examination Platforms

• AI Resume Builder

• Learning Management Systems

Track B
~35% of students

Data Science & AI/ML

Applied ML & Analytics

• Performance Prediction

• Placement Analytics

• LLM Evaluation

• Recommendation Systems

Track C
~15% of students

Research-Oriented

Advanced AI Research

• Prompt-Injection Robustness

• Multi-Agent Systems

• AI Safety Evaluation

• Mechanistic Interpretability

Project Domains (60 Topics)

Domain A

Artificial Intelligence & Machine Learning

15 projects
Domain B

Data Science & Visualization

10 projects
Domain C

Python Packages & Full-Stack Tooling

10 projects
Domain D

MERN / Full-Stack Systems

10 projects
Domain E

Cybersecurity + AI

8 projects
Domain F

Advanced Research-Oriented

7 projects

Evaluation Scheme

The defining feature of this rubric is that Testing & Evaluation (15 marks) is weighted independently and substantially, separate from Implementation. A project without proper evaluation is considered incomplete regardless of implementation quality.

Problem Definition10
Literature Review10
Architecture & Design10
Implementation20
AI/ML/Data Component15
Testing & Evaluation15
Innovation10
Documentation/Paper5
Presentation/Demo5

⚠️ Important: A submission reporting only final performance without baseline comparison, unseen-data validation, and error-category breakdown is considered incomplete under Testing & Evaluation.

Program Requirements

Team Formation

  • • Teams must have 1-4 members
  • • One member designated as team leader
  • • All members must be registered
  • • Track assignment based on ability and interest

Weekly Commitment

  • • Weeks 1-2: 3-4 hours (Foundation)
  • • Weeks 6-13: Higher (Sprint-dependent)
  • • Weeks 14-16: Moderate (Documentation)
  • • All sessions after college hours

Deliverables

  • • Working software artifact
  • • Research paper documenting evaluation
  • • GitHub repository with documentation
  • • Presentation and demonstration
  • • Poster presentation

Foundation Phase (Weeks 1-2)

  • • Git/GitHub workflow training
  • • Python and JavaScript refresher
  • • Research methodology introduction
  • • Literature search techniques
  • • Project proposal submission

Ready to Start Your Research Journey?

This program moves beyond conventional "mini projects" into a research-oriented engineering exercise where every project produces both a working artifact and a documented evaluation against measurable baselines.

Prepared by: Prof. Nandhan K & Prof. Md Usama Thangalvadi

Department of Computer Science & Engineering

Gandhi Engineering College (Autonomous), Bhubaneswar