AI Enabled Test Bed Generator

Published: 21 May 2024| Version 1 | DOI: 10.17632/j78szr5p99.1
Soham Som, Kailas PATIL, prawit chumchu


Background: It might take a lot of time and effort to create high-quality test beds for software testing that accurately reflect real-world events. Even with automated tools, it might be difficult to guarantee the inclusion of pertinent code smells or accurate data; manual generation frequently calls for a large amount of work. Problem Statement: Comprehensive software testing depends on effectively creating test beds that faithfully capture the subtleties and complexities of real systems, including code smells and plausible situations. Such features may be difficult to add to existing systems, which might result in inadequate test coverage and efficacy. Solution Provided: Code Smells Database: A database containing high-quality code smell samples used for static content generation. Static Content Generator: A component that generates test beds using samples from the Code Smells Database. ChatGPT API: An external API for ChatGPT that is used to generate test beds dynamically based on natural language input. Ollama Engine: An in-house large language model (LLM) engine that can be triggered to generate test beds through specific commands (e.g., "Ollama"). Ollama REST API Client: A client that interacts with Ollama's REST APIs to trigger test bed generation. You can find the updated code here. Github URL :



Software Engineering, Software Tool, Software Testing, Testbed