🤖 QAgent – Automated QA Test Case Generation Tool

QAgent is an intelligent web-based tool that automates the generation of QA test cases from documents and screenshots using the power of AI. Just upload a Software Requirements Specification (SRS) file or a UI image, click a button, and get a well-formatted Excel file with detailed test cases — ready for your QA team to use.

---

🚀 What It Does

QAgent reads through:

• 📄 SRS documents (.docx or .pdf)
• 🖼️ UI screenshots or mockups (.png, .jpg, .jpeg)

Then:

1. Extracts text and UI components using OCR and YOLO (object detection).
2. Feeds the structured information to the OpenAI GPT-4 model.
3. Automatically generates detailed test cases in Excel-ready format (Test Case ID, Description, Preconditions, Steps, Expected Result, Status, Comments).

---

🧠 Why It Matters

In real-world QA workflows, teams spend hours manually writing test cases from functional specs and design screens. This process is:

• Time-consuming
• Prone to human error
• Tedious and repetitive

QAgent changes the game by using AI to automate this process, saving QA engineers hours of manual work while improving coverage and consistency.

---

🛠️ Technologies Used

🔹 Frontend (React + Next.js)

• React 19 with Next.js 15
• Tailwind CSS 4 for fast, responsive UI design
• Radix UI + Lucide Icons for beautiful accessible components
• React Hot Toast for notifications

🔹 Backend (Python + Flask)

• Flask API with file handling and response routing
• OpenAI GPT-4 API to generate test cases from structured content
• Pytesseract for OCR text extraction from images
• YOLOv8 (via Ultralytics) to detect UI components like buttons, inputs, dropdowns
• fitz (PyMuPDF) for reading PDF files
• docx (python-docx) for reading Word documents
• Pandas to format and write Excel output
• OpenCV and PIL for image processing
• ExcelWriter (xlsxwriter) to save the final output

---

📁 Project Structure

QAgent/ │ ├── app.py # Flask backend ├── requirements.txt # Backend dependencies ├── frontend/ # Next.js + React frontend │ ├── pages/ │ ├── components/ │ └── ... ├── Input files/ # Uploaded SRS or image files ├── Generated Output/ # Auto-generated Excel files ├── Input images/ # Temporary image extractions └── .gitignore

---

⚙️ How It Works (Under the Hood)

1. User uploads file from frontend

   • Frontend sends a POST request to /qa_testcases with the uploaded files.

2. Backend processes files

   • PDF/DOCX → text via fitz and docx
   • Images → text via pytesseract, components via YOLO

3. Combined data is fed to GPT-4

   • Prompted with a structured instruction to generate test cases

4. Test cases parsed and saved

   • Output saved as output.xlsx in temp directory

5. User downloads result

   • Frontend provides direct link to /output.xlsx to download test cases

---

🌍 Real-World Use Cases

Scenario	Impact
🔧 Manual QA Teams	Automatically generate dozens of test cases from lengthy requirement docs
🧪 Agile Startups	Speed up sprint testing with quick turnaround
🧩 UI/UX Designers	Validate designs by generating test expectations from screenshots
🏢 Enterprises	Standardize test case formatting and reduce errors

---

📌 Skills Demonstrated

• End-to-end Full Stack Development
• AI Integration with OpenAI GPT-4 API
• Working with unstructured documents and OCR
• Object detection using YOLOv8
• Excel automation using Python
• Clean and responsive UI/UX using modern React stack
• Git, GitHub project management

---

📜 License

This project is open-source and free to use under the MIT License.

---

🙌 Author

Built by Giri Ratna Sai Gummadi
🔗 LinkedIn | 🌐 GitHub