Artificial intelligence is transforming businesses’ operations, and software testing is no exception. As organisations increasingly adopt AI tools, the impact on efficiency, accuracy, and scalability is profound. According to the CompTIA Community, AI is projected to create 12 million more jobs than it will replace, underscoring its role as a catalyst for workforce evolution rather than disruption. Recent figures from IBM reveal that 30 per cent of IT professionals are leveraging AI and automation to save time, highlighting its practical applications in streamlining operations.
With 35 per cent of businesses integrating AI into their workflows, its influence spans diverse functions, from administration and marketing to sales and IT. AI enables enterprises to overcome long-standing challenges in software testing, such as resource constraints and increasing complexity, by introducing intelligent automation. As these technologies mature, they are reshaping the testing landscape, reducing bottlenecks, and setting new benchmarks for quality and speed.
The evolution of AI in software testing
Generative AI has been present in the market for some time, but what is happening now is its evolution into a truly transformative force. “The ability of generative AI to perform tasks autonomously and make decisions in a way that mirrors human professionals is the most significant development,” Tal Barmeir, serial entrepreneur and co-founder of BlinqIO, says. “It has matured to the point where instructions can be given in plain English, similar to briefing a new employee, and the AI can interpret, implement, and deliver results at a high level of quality.”
This maturity allows for creating virtual brains that act as software engineers, testers, DevOps professionals, or IT administrators. These virtual entities can be scaled infinitely. Whether an organisation needs one, ten, or ten thousand of them is simply a matter of computing power. This is a dramatic departure from the traditional human resources and infrastructure limitations. Instead of provisioning physical servers, companies are deploying virtual engineers capable of performing highly complex tasks with precision, scalability, and immediacy.
From mobile automation to AI-driven testing
Barmeir’s career has consistently been in the software industry, particularly software testing. When smartphones began dominating the market, she developed the first test automation platform to handle the vast array of mobile configurations and devices. At the time, this was revolutionary. “Now, my focus has shifted to leveraging generative AI to create AI-powered test engineers,” she adds. “Testing is a critical component of software development, and bottlenecks, talent shortages, and inefficiencies have always plagued it. Using AI to address these issues has been exciting for years, but the technology was not ready until recently.”
For many startups, AI is the driving force, but for Barmeir, it is simply a means to an end. “Alongside my co-founder, Guy Arieli, I have founded three companies, all of which were bootstrapped and two of which were acquired,” she continues. “When you bootstrap a company, there is no room for error in identifying and delivering business value. You must generate revenue from day one to survive, so your focus has to be laser sharp.”
The company’s AI-powered test engineers address enterprises’ significant challenges in software testing. Finding and retaining skilled test automation engineers is incredibly difficult; without them, organisations struggle to release software on time and at a high quality. This technology solves the problem by automating the testing process and eliminating the need for large teams of human testers.
Building AI models for enterprise-grade testing
AI-powered test engineers were an ambition long before generative AI reached its current capabilities. Machine learning, AI, and now generative AI have all been part of the broader vision, but early iterations required significant human intervention to function effectively. “Around 18 months ago, we identified that generative AI had reached a tipping point,” she explains. “The models had matured to the level where they could deliver results equivalent to those of a human test engineer regarding quality and accuracy. The speed at which generative AI is advancing is unprecedented. In the 25 to 30 years I have spent in the software testing industry, I have never seen technology evolve so rapidly.”
BlinqIO uses a combination of general-purpose AI engines and proprietary models to meet the specific demands of software testing. Large language models from OpenAI and Google play a role in their systems, but they are insufficient for the complexity of testing tasks. “Software testing requires understanding a fully digital ecosystem, multi-application environments, complex user interfaces, and dynamic workflows,” Barmeir explains. “Large language models excel at text-based tasks but fall short when comprehending and interacting with these digital environments. We have trained our own open-source models specifically optimised for software testing. These models are designed to handle the intricacies of testing tasks, from comprehending a website or application to interacting with its elements.”
A retrieval-augmented generation (RAG) algorithm enhances the AI’s ability to contextualise its tasks. By integrating data from enterprise systems like JIRA and publicly available information, BlinqIO’s AI test engineers can understand the context of the application they are testing, resulting in far better performance.
Automating accuracy in high-risk industries
The software testing market has long struggled with bottlenecks and inefficiencies. Test automation, in particular, suffers from a shortage of skilled professionals. Many talented programmers move on to other roles in research and development, leaving the testing field with a significant talent gap. This creates a backlog of manual testing tasks, which slows down the entire software development lifecycle.
“Our AI-powered test engineers solve this problem by automating the creation and maintenance of test automation code,” Barmeir says. “They can be operated by manual testers who have no coding expertise. The AI can autonomously adapt to changes in user interfaces, ensuring that tests remain accurate and up-to-date without human intervention. If a menu item is relocated in an application, the AI can identify the change, update the test code, and proceed without requiring hours of human analysis and rework.”
Rigorous quality assurance is essential for industries such as banking, finance, healthcare, and aviation. Errors in financial applications, for example, can have serious consequences, while mistakes in healthcare software could impact patient outcomes. “These industries are highly sensitive to quality and cannot afford the risks associated with manual or inefficient testing processes,” Barmeir says. “Our technology is designed for environments where precision and reliability are paramount.”
Generative AI requires significant computing power, much like the human brain needs energy. While BlinqIO’s basic offering is cloud-based, it provides on-premises deployments for clients in highly secure industries like banking. “Training our proprietary models involves focusing on specific capabilities related to testing tasks,” Barmeir adds. “This targeted approach allows us to optimise our models without the vast computational resources required for general-purpose AI systems like OpenAI.”
The future of AI-powered testing
BlinqIO’s vision is to revolutionise the software testing industry by eliminating manual testing entirely. “We aim to provide fully automated testing across all platforms, languages, and test types, including functional, API, load, and performance testing. Our AI test engineers support multilingual testing and can seamlessly transition between mobile and web platforms. Ultimately, we want to create a single test requirement process where AI can perform tests across multiple platforms and languages with the same ease as a human engineer.”
The biggest challenges for generative AI remain trust and security. Enterprises are cautious about adoption due to concerns over data security and the potential for job displacement. However, Barmeir sees this as an opportunity rather than a threat. “Generative AI is not about replacing people but about enhancing their productivity,” Barmeir concludes. “Just as computers did not eliminate jobs but created new opportunities, AI will allow humans to focus on more complex and creative tasks. If managed responsibly, it can drive significant improvements in productivity, efficiency, and innovation across the board.”
