# 🧑‍🏫 Teacher Copilot: AI Feedback Engine for Teachers

**Teacher Copilot** is an intelligent, AI-powered assistant built to evaluate student coursework efficiently. By automating repetitive feedback and identifying class-wide learning gaps, it significantly reduces grading time while ensuring teachers remain in complete control of the final output.

## 📌 The Problem

Educators spend a disproportionate amount of time checking assignments, identifying common mistakes, and writing repetitive feedback. This administrative burden subtracts from valuable time that could be dedicated to lesson planning, one-on-one student interaction, and curriculum improvement.

## 💡 The Scenario & Expected Solution

**Scenario:** A teacher uploads a batch of student submissions. The Teacher Copilot system analyzes these submissions to generate constructive feedback and actionable insights.

**Expected Solution:** We have built an AI-driven pipeline that evaluates student work accurately, detects common misconceptions, and presents the data on an intuitive dashboard. The human-in-the-loop design ensures that teachers can review and edit all AI suggestions before finalizing.

## ✨ Core Features

- **📥 Seamless Input Processing:** Upload student answers in bulk via text, PDF, or document formats.
- **🤖 Contextual AI Feedback:** Generates personalized, constructive feedback for individual students based on a provided rubric.
- **📊 Common Mistake Detection:** Analyzes the entire batch of submissions to highlight recurring errors and knowledge gaps across the classroom.
- **🎛️ Interactive Teacher Dashboard:** A centralized control hub where educators can review, override, or approve AI-generated feedback.
- **📤 Clear & Exportable Output:** Generates clean, readable reports ready to be shared with students or integrated into an LMS.

## 🛠️ Tech Stack

* **Backend:** Python, FastAPI / Flask
* **Frontend/Dashboard:** Streamlit / React
* **AI & NLP:** Large Language Models (OpenAI API / Google Gemini), LangChain
* **Data Processing:** Pandas, PyPDF2

## 🚀 Getting Started

Follow these instructions to set up the project locally on your machine.

### Prerequisites

* Python 3.8+
* API Key from your LLM provider (e.g., OpenAI or Google Gemini)
* Git

### Installation & Setup

1. **Clone the repository**
   ```bash
   git clone [https://github.com/aashuutosh/teacher-copilot.git](https://github.com/aashuutosh/teacher-copilot.git)
   cd teacher-copilot

2. Initialize Environment & Install Dependencies

# Create virtual environment
python -m venv venv

# Activate on Windows:
venv\Scripts\activate
# OR Activate on macOS/Linux:
source venv/bin/activate

# Install requirements
pip install -r requirements.txt

3. Configure Environment Variables Create a .env file in the root directory and add your API key:

LLM_API_KEY=your_secret_api_key_here

4. Run the Application

# If using Streamlit:
streamlit run app.py

# If running standard Python backend:
python main.py

🔄 Workflow Architecture

1. Data Ingestion: System reads the assignment prompt, grading rubric, and student files.
2. AI Processing: Submissions are tokenized and sent to the LLM alongside the rubric instructions.
3. Insight Extraction: The model evaluates the text, scores it, and extracts common structural errors.
4. Dashboard Render: Data is passed to the frontend for the teacher to review.
5. Finalization: Approved feedback is exported into a standardized format.

🔮 Future Enhancements

• LMS Integration: Direct plugins for Google Classroom, Canvas, or Moodle.
• Handwriting Recognition (OCR): Support for scanned handwritten assignments.
• Longitudinal Tracking: Track a student's progress and mistake patterns over an entire semester.

👨‍💻 Author

Ashutosh Pratap Singh

• GitHub: @aashuutosh

If you find this project helpful, please consider giving it a ⭐ on GitHub!