Artificial Intelligence in Software Quality and Testing: Exploring Machine Learning & Large Language Models

Artificial Intelligence is great at mimicking human intelligence bringing in capabilities of visual processing, image recognition, data analysis and generation, natural language processing and decision making. Software quality analysis and testing requires many of these same human intelligence skills, these skills can be augmented by AI to further improve end to end software quality.

AI can bring significant improvements to overall software test planning, design, execution, code reviews, automation, security, metrics, tooling, resource and risk management. With the emergence of low-code or no-code software testing tools using machine learning algorithms and models, it’s evident that software testing is becoming more robust, faster, resilient and efficient. Machine Learning can train and build models that can help identify common defect areas and failures in software code base on several development cycles, the models are constantly learning and evolving, getting better at identifying and predicting issues and defects even during the active software development.

The lines between software development and testing have never been so fuzzy with the insertion of AI technologies in overall software development. With the emergence of Generative AI and Large Language Models recently, testing is becoming more intelligent and creative. This paper explores few applications of Machine Learning training and inference in modern software testing, the potential of generative AI, the impact on the software industry and some of the challenges with adopting AI for testing.


Amith PullaAmith Pulla

Amith Pulla is a Software Program Manager at Intel Corp, currently working at Intel site in Hillsboro, Oregon. Amith works with engineering teams developing full-stack software for SoC platforms for Server and Client products. Amith works with software enabling and validation teams including BIOS, Firmware, Operating Systems, Compilers, Libraries and Frameworks. Amith works on platform software enabling plans in pre-silicon and post-silicon to ensure software stack meets the quality criteria for the product launch.

Over the last 21 years, Amith has been involved in software enabling, testing strategies and processes for applications from sales, marketing to data analytics and system software. Amith also worked in market research and intelligence space for data center and server market. Amith has worked extensively on projects involving multiple platforms and complex architectures. Amith worked on developing test methodologies and techniques to meet the business needs and schedules. As part of his QA lead role for the first 8 years of his career, Amith focused on improving and refining QA processes and standards for efficient software testing in agile environments. Amith also took on ScrumMaster and agile coach roles working with agile development teams.

Amith has an M.S. in CIS from the New Jersey Institute of Technology and MBA from University of Illinois, Urbana Champaign. Amith got his CSTE and PMP certifications in 2006, CSM certification in 2012.