🚀 SyncETA: "Understand like a human, test like a machine"

AI-based Web Browser Action Recording and Automated Test Execution System

SyncETA is an innovative test automation solution that records web browser-based user actions with Selenium, automatically generates natural language-based test cases utilizing LLMs (Large Language Models), and automatically executes them through Playwright MCP. It aims to maximize the efficiency of QA work and continuously improve software quality.

View Product Introduction

💡 Problems SyncETA Aims to Solve

Despite test automation being essential in the existing software development and deployment process, there were the following difficulties:

• High implementation costs and maintenance burden: Significant QA resources are consumed in implementing and maintaining test automation solutions.
• Manual-centric inefficiency: Test case writing is human-centric, resulting in low repeatability and reusability, and a high probability of human error.
• Increased regression testing burden: The burden of repetitive regression testing is exacerbated due to frequent release cycle shortening.
• Development knowledge required: Accessibility for non-technical users (QA teams) was low because development knowledge was required to write scripts for test automation.

SyncETA solves these problems, providing an integrated solution where all users, including QA engineers, can easily build test automation, reduce maintenance costs, and maximize the efficiency and quality of regression testing.

🛠️ SyncETA's Core Technologies and Solutions

SyncETA solves problems and provides an innovative test automation experience through the following key modules.

1. SyncETA Recorder (Selenium-based Browser Action Recording Module):

• Captures the actual web browser actions of the user (clicks, inputs, scrolls, page transitions, etc.) in real-time and automatically generates and saves them as test scenarios in a JSON/YAML-based format.
• Information such as URL, XPath, Class, ID, etc., is recorded together during recording, and the recorded scenarios can be classified, searched, and version-controlled through the management function.
• Provides a user-friendly screen such as inputting the recording web URL and free web browser selection and size setting, so anyone can easily record.

2. TestCase Generator (LLM-based Natural Language Test Case Auto-Generation Module):

• The recorded scenario (JSON) is analyzed by a 7B or open-source LLM and converted into a natural language-based Excel test case including Step, Input, Expected Result, etc.
• During this process, context learning about in-house terminology and UI elements, and format customizing according to QA standards are supported.
• It can be provided in the form of reusable templates.

3. Test Executor (Playwright MCP Interlocking Auto Test Execution Module):

• For test cases approved by QA, an auto-execution script is generated through Playwright MCP, and supports parallel test execution in a multi-browser automation environment. It supports compatibility with various browsers such as Chrome, Firefox, and Edge, and the addition of the Safari browser has also been requested.
• Provides a function for the AI to directly execute Excel test cases created or possessed by QA directly without event recording, maximizing the flexibility of test execution. The AI analyzes fields such as 'Step', 'Input', and 'Expected Result' in the Excel file, performs the action in the browser, and verifies the result.
• Provides an execution result report (pass/fail log, screenshots, etc.), and failed test cases are automatically logged.
• Identification is possible when an error occurs during scenario execution, and provides user-friendly error messages and diagnostic features such as error record screenshots and providing the recorded screen (video) at the time of the error when clicking the error on the dashboard.

4. QA Feedback Loop (Quality Improvement Loop Module):

• QA engineers can directly review, modify, and evaluate the generated Excel test cases, and the modified data is uploaded to the SyncETA system and reflected in the LLM's re-training (Fine-tuning or RAG).
• Secures LLM training data based on QA's modifications and continuously improves test case generation quality through periodic re-training. This increases the reliability of auto-generated test cases and improves test coverage. It is also possible to analyze feedback history by QA and track improvement details.

✨ SyncETA's Key Features and Characteristics

• Dashboard/Management System: Provides a dashboard for system monitoring such as scenario registration status, execution status, and error status. You can check the video in the execution history, and it helps with error analysis by providing screenshots of error records and recorded videos of the screen when errors occur. Provides precise reports and analysis, and can provide custom reports tailored to customer requirements.
• Data Set Management Feature: Inputs and manages data in a tabular format, supporting the addition, modification, and deletion of variables. You can write directly by calling variable values set in the scenario, or automatically generate mockup data using AI and conduct tests with various values.
• Scenario Management and Execution Options:
  • Record: The basic unit that records user events and DOM information.
  • Scenario: A core component that implements test scenarios by systematically integrating record information.
  • Collection: A feature that selects multiple scenarios and executes them sequentially, guaranteeing a consistent workflow by flexibly combining various scenarios. It is used when simultaneously executing one scenario across multiple browsers.
  • Story: A feature that freely connects scenarios through a Flow Chart interface and dynamically executes the next scenario based on verification results.
  • Dynamic Test Chaining: Enables flexible testing through sequential scenario execution based on conditions.
  • Provides various features to facilitate test management, such as initialization upon re-execution, scenario severity setting, and scenario history filtering.
• Screen Verification: Additional data verification is possible after capturing the screen for each event through AI, and when UI changes are detected, it can automatically process the test as a failure.
• Real-time Communication: Enables checking real-time event actions and DOM information through socket communication.
• Security and Authentication Processing: Supports automatic login, cookie and session management, HTTP authentication bypass, VPN/Proxy settings, etc.
• CI/CD Integration: Seamlessly integrates test automation into existing development/deployment processes by interworking with various external tools/systems such as Jenkins, Git, Azure, DevOps, etc.
• Code/Non-code-based Test Design: Even users lacking development knowledge can easily build test automation with an intuitive GUI and keyword-based non-code methods, and it also provides a script (JavaScript)-based code method for experts.
• API Test Support: Provides an all-in-one test platform by supporting not only web tests but also API tests.

🎯 SyncETA's Expected Effects

SyncETA contributes to securing business agility and competitiveness by providing the following innovative effects:

• Reduce test case writing time by 80%: Drastically shortens test case writing time compared to manual work through Selenium recording and LLM-based automatic generation.
• Streamline QA Resources: Repetitive testing tasks are automated, increasing QA work efficiency and reducing QA resources by more than 30%.
• Improve Test Quality and Reliability: Through AI-based test case generation and the QA feedback loop, test accuracy and coverage are improved, human errors are prevented, and consistent quality is maintained.
• Fast Release and Time-to-Market: Shortens the development cycle through regression test automation and accelerates time-to-market with rapid feedback.
• Test Automation without Writing Code: Users lacking development knowledge can also easily build test automation with an intuitive GUI and keyword-based non-code methods.
• Provide a Flexible Test Environment: Supports various execution options and customization such as cross-browser compatibility, parallel execution, and dynamic chaining.
• User-friendly Error Identification: Facilitates error analysis by providing user-friendly error messages when an error occurs, recording screenshots, and providing recorded screens when an error occurs.

🌐 Target Audience and Deployment Format

SyncETA can be deployed as a Docker-based installation (On-premise) or in a Cloud SaaS format, and depending on security requirements, the LLM model can be executed on-premise.

• Target: SaaS web service QA teams, large commerce platforms, enterprises requiring regression test automation, web application development teams, QA engineering teams, DevOps teams, AI/ML teams, etc.

🏆 SyncETA's Differentiated Strengths

Based on the commercialization potential of rapidly rising open-source LLMs and the explosive demand for QA automation, SyncETA provides a quality improvement model based on an LLM learning feedback loop that competing solutions do not have. Furthermore, it emphasizes 'Customization in its DNA', which can flexibly adjust and expand test environments and features according to customer's specific requirements, aiming to become a strategic tool contributing to enterprise digital transformation and strengthening competitiveness.

---