An Intelligent Tool for End-to-End Automated Testing
Modern software testing is time-consuming, repetitive, and costly. Manual QA struggles to keep up with rapid development cycles, leading to delayed deployments and potential bugs slipping into production.
Our AI-Powered Test Automation Agent addresses this by providing an end-to-end autonomous testing pipeline. From analyzing Figma designs to generating structured test cases, executing them using Selenium, and seamlessly integrating into CI/CD pipelines — our system eliminates manual bottlenecks and accelerates reliable software delivery.
This project was developed during HackNuThon 6 by Team Incognito.
- Jainil Patel
- Raj Makwana
- Krish Chothani
- Het Patel
- AI-Driven Test Case Generation: Extracts functional & UI test cases directly from Figma designs.
- One-Click Test Automation Agent: From input design to automated execution, no manual steps needed.
- Dynamic Web Scraping: Adapts Selenium scripts by fetching relevant HTML element IDs and classes.
- Parallel Execution: Multi-threaded test execution reduces overall time.
- CI/CD Integration: Seamlessly integrates into GitLab pipelines for continuous testing.
- Structured Logging & Reporting: Generates JSON-based logs for reproducibility and debugging.
- Scalable AI Backend: Uses multiple API keys and load balancing to overcome API rate limits.
- Frontend: React.js, Node.js
- Backend: Python, Shell Scripting
- Automation Framework: Selenium
- Database: MongoDB
- CI/CD: GitLab
- AI Model: Google Gemini API
The problem statement challenged us to develop an intelligent AI Agent capable of automating the testing process.
- High costs due to manual testers.
- Limited adaptability of Selenium scripts across different web architectures.
- Delays caused by sequential test execution.
- API rate limits restricting speed of test case generation.
- Lack of structured integration with CI/CD pipelines.
✅ Dynamic Script Adaptation – Extract HTML elements dynamically to make Selenium scripts robust across websites.
✅ Context-Aware Error Detection – Structured validation checks to handle diverse website architectures.
✅ Optimized Prompt Engineering – Strictly structured outputs improve accuracy of AI-generated test cases.
✅ Multi-threaded Parallel Execution – Multiple API keys with threading overcome Gemini API rate limits.
✅ CI/CD Automation – YAML-based GitLab pipeline ensures fully automated execution with every commit.
- Requirement Extraction – Analyze Figma designs for UI-based test cases.
- AI-Generated Test Cases – Convert requirements into structured JSON format.
- Web Scraping & Mapping – Fetch relevant IDs, classes, and elements dynamically.
- AI-Powered Script Generation – Generate Selenium test scripts.
- Automated Execution – Scripts executed via CI/CD pipelines.
- Reporting – Results logged in JSON format for debugging & monitoring.
📌 Architecture Diagram:
| Challenge | Our Solution |
|---|---|
| Selenium scripts not adaptable to all websites | Used dynamic web scraping for element detection |
| Error detection varies across websites | Built context-aware validation checks |
| API rate limits on Gemini | Multi-threading + multiple API keys |
| Accuracy of generated test cases | Optimized prompt engineering |
| Sequential execution slows testing | Parallel test execution |
| CI/CD integration | Custom YAML GitLab pipeline |
- ⚡ Instant Test Execution – Developers test code instantly.
- 🚀 Faster Deployments – Reduced deployment time & faster go-to-market.
- 💸 Cost Reduction – Eliminates the need for manual testers.
- 🎯 Improved Accuracy – Reliable and adaptable test case generation.
git clone https://github.com/raj-mistry-01/PipeTest-Frontend.git
cd PipeTest-Frontend
npm install
npm run devAI-Powered Test Automation Backend
- Project Demo Video: LINK
- Project Screenshots:
