Enterprises often struggle with efficient testing and maintenance of Guidewire Insurance Suite applications using conventional methods. Such testing approaches demand significant time, resources, and skillsets making it difficult to keep pace with frequent updates.
To overcome these challenges, Infosys offers an automation solution powered by generative artificial intelligence (GenAI), which optimizes and accelerates Guidewire quality engineering (QE) processes.
Infosys leverages its AI platform to automate key stages of the Guidewire testing lifecycle. The solution combines GenAI with a structured workflow to automate the generation of key testing assets.
Leverages LLM capabilities and embedded, context-sensitive learning extracted from client-specific knowledge databases, such as user guides, user stories, test cases, Guidewire testing automation frameworks, automation scripts, and object repositories
Provides a set of predefined prompts
Facilitates smooth integration with test management suites and code repositories, including Jira, QTest, and GitHub
Incorporates quality engineer reviews to enhance accuracy and reduce operational overheads
Enables the creation of custom workflows by integrating components such as:
Connectors: Jira Pull/Push and ADO Pull/Push, among others
Chains: LLM chain, retrieval chain, and more
Utilities and tools: Output parser, unstructured file loader
Fast-tracked creation of test scenarios, test cases, and automation scripts
Enhanced test quality with comprehensive coverage
Reduced testing costs and manual effort
Increased efficiency in Guidewire application testing
Accelerated time to market for Guidewire implementations and updates
Analyzes business requirements and generates user stories by referencing Guidewire user guides for applications such as PolicyCenter, BillingCenter, and ClaimCenter, along with custom implementation details
Creates detailed test cases automatically from generated user stories
Harnesses GenAI to generate automation scripts based on test cases using frameworks or scripting languages such as Python, Java, and Selenium
Executes automation scripts by implementing web agents
Embeds contextual data corresponding to Guidewire portal documents, including configuration files, XPath repositories, and testing frameworks
Quality engineers have the option to manually review GenAI user stories, test cases, and automation scripts to ensure quality. Until the model is thoroughly trained for client-specific context generation, we recommend human-in-the-loop (HITL) intervention to fine-tune training for enhanced model accuracy.