When manual testing was the norm, it was common for software companies to hire a full-time Quality Assurance (QA) team to develop a collection of “test plans” to ensure software projects’ features were performing as expected. This team would manually run checklists each time an update was introduced to a software project, and then pass on the test plan results to the engineering team to address any issues.
This process was slow, expensive, and error prone. Moreover, testing the web routes that are crucial to a business and identifying those most likely to fail or those most often used by end users can be challenging.
Automated testing is a process that guarantees quality at all stages of development, ensuring that software works correctly before deployment and that updates do not introduce bugs. This greatly increases the efficiency of QA and Quality Control (QC) teams by allowing them to focus on more critical tasks.
In pursuit of efficiency
At Amaris Consulting, the way we work is constantly evolving to ensure that our clients and partners achieve the best possible results. This is how our QA team came up with AI4QA, a cutting-edge AI system.
After successfully completing a complex banking and finance project, our QA team decided to rethink our automated testing process. The team wanted to continue seeing high quality results but find a simpler and faster way to do so. The goal was to build a system that would perform automated code generation and analysis: to reduce project duration and allow automation engineers to spend more time gathering the data needed to make informed decisions. The team thus developed AI4QA: a system that analyzes and semantically understands web data.
The magic behind the curtain
AI4QA extracts data in a way which mimics how users browse the web. This method uses the latest machine learning classification techniques to enable the software to analyze and classify data and solve problems on its own.
This new system will revolutionize the way QA engineers test their applications. AI4QA users will always be in control of the testing process. However, the role of the QA team members will change radically as they will no longer build the code ̶ instead they will decide whether the machine-built code is relevant or not. The process is thus simplified. Since building code will be controlled by the machine, automated test cases can be built in a few weeks rather than months. This system will allow automation engineers to spend their time making decisions instead of coding.
At Amaris Consulting, we are dedicated to helping you solve your toughest engineering challenges. Discover how our engineering solutions can benefit you.
|Project title (original)||Desarrollo de sistema basado en Inteligencia Artificial como potenciador de diseño de pruebas de software|
|Project title (in English)||Development of an AI-based system for enhanced design of software testing cases|
|Project type||Research and Development|
|Co-funding body||IVACE – Valencian Institute of Business Competitiveness|
|Co-funding Program||I+D EN COOPERACIÓN (PIDCOP-CV) 2021|
|Implementation area||Valencia, Spain|