# LabCode Fixer & Viva Coach

> Source: <https://dev.to/varshininamrathamopuri_51/labcode-fixer-viva-coach-4ac0>
> Published: 2026-10-03 09:29:59+00:00

**Don't just fix the bug. Learn how to defend the code.**

LabCode Fixer & Viva Coach is an AI-powered debugging and viva preparation tool built for me and a fellow college student who work with programming assignments and laboratory exercises.

While debugging a program, fixing the error is only one part of the problem. During a laboratory viva, a student may also be asked:

I wanted to build something that helps with more than just getting code to run.

The idea is simple:

**Debug → Understand → Correct → Defend**

The student provides:

The application then provides:

The current version supports:

The goal is not just to help students fix errors, but to help them understand the fix and prepare to explain their code during a viva.

The demo shows the complete workflow:

**Enter code → Enter error → Analyze → Understand the error → See corrected code → Practice viva questions**

**Demo video:**

The demo is approximately 1 minute and 31 seconds long and focuses on the working application.

The complete source code is available on GitHub:

LabCode Fixer & Viva Coach is a student-focused AI debugging and viva preparation tool.

It helps students understand programming errors instead of simply copying a solution. The application explains why the code failed, provides a corrected version, and generates three technical viva questions based on the student's actual code.

Students often use AI tools to fix compiler or runtime errors, but they may still struggle to explain the solution during a laboratory viva.

A professor may ask:

LabCode Fixer & Viva Coach is designed to help students understand their errors and prepare for these questions.

The application…

The repository contains:

`backend/main.py`` backend/requirements.txt``frontend/app.py`` README.md``LICENSE``.gitignore`
The Python virtual environment and downloaded AI model files are not included in the repository.

The project uses a simple architecture where each component has a clear responsibility.

```
text
Student
   |
   v
Streamlit Frontend
   |
   v
FastAPI Backend
   |
   v
Ollama
   |
   v
Gemma 2B
   |
   v
AI Analysis
   |
   v
Streamlit Frontend
```