Automated Test Generation and CI/CD Construction Support Services
DXE -Digital eXploring Engagement business-
*Note: This service is provided exclusively within Japan. We do not conduct business or provide support overseas.
Providing End-to-End Support via Generative AI and Automated Environments to Address Escalating Testing Workloads and Quality Demands
- Target Industry : Automotive / General Manufacturing
Problem
Are manual tasks lurking behind your test automation?
- Manually updating specifications and test cases for every specification change imposes a heavy workload to track modifications.
- The massive workload makes full regression testing difficult, while test prioritization relies on individual discretion, leaving constant anxiety over missed defects.
- Compliance with safety standards such as IEC 61508 cannot be left entirely to AI, making the establishment of human review systems a heavy burden on front-line teams.
Solution
Linking AI-Driven Creation, Execution, and Remediation Based on Existing Assets: Toward a Development Environment That Reinvests Engineering Hours into Quality Improvement
Starting from existing specifications and test assets, we deliver solutions that achieve both quality and speed by linking AI generation with CI/CD automated execution. Going beyond merely providing automation tools, we align closely with on-site safety requirements and review processes, providing end-to-end partner support—from planning an optimal collaborative framework between AI, humans, and tools to environment setup and operational adoption.
Features
-
Features 1
AI-Driven Automated Test Generation
The AI accurately understands specifications created from "unstructured data" such as images, Word, and PDF files, automatically generating high-precision test data.
-
Features 2
Upgrading the CI/CD Execution Environment
• Streamlining builds through partial compilation in a build optimization environment
• Automated generation of stub functions and dummy variables using test data generation AI
• Eliminating dependence on target hardware and achieving automated execution by building a test automation environment
Two Key Points for Achieving Automated Test Generation and CI/CD
-
AI analyzes specifications and existing test assets, formatting them into an intermediate structure for human review and approval. Starting from this intermediate format, test data is generated and seamlessly integrated with CI/CD for automated execution, evaluation, and reporting.
-
1． Implementation of "Quality Gates" to Prevent Low-Quality Outputs
Rather than having AI directly convert specifications into test data upon understanding them, it generates test specifications in an "intermediate format" (Markdown/CSV) with explicitly defined test objectives and conditions. By inserting quality gates that enable human review as well as coverage and failure result evaluations at this intermediate stage, both explainability and quality are achieved.
-
2． CI/CD Integration Plan and Pipelining
In CI/CD, the process from change detection to regression completion is structured as a phased pipeline. While generation and execution are automated, approval steps are included where human judgment is required to ensure auditability.
Phase Design
-
Phased Design for Step-by-Step Improvements in Automation Rate, Accuracy, and Coverage
In the initial stage, we narrow the scope to implement automated generation and CI/CD integration in a measurable way. While verifying effectiveness at each phase, we ultimately advance to a closed-loop model that feeds execution results back into learning.
Co-creating unprecedented evolution through unstoppable technology.
Please feel free to contact us for details on our products and services or to request materials.