Innovative Methods in Automated Accessibility Testing

Available online www.ejaet.com European Journal of Advances in Engineering and Technology, 2025, 12(11):7-13 Research Article ISSN: 2394 - 658X 7 Innovative Methods in Automated Accessibility Testing Santhosh Kumar Jayachandran Independent Researcher, USA _____________________________________________________________________________________________ ABSTRACT Automated accessibility testing is developing at a rapid pace, and new methodologies are being developed to improve digital inclusion. This article describes cutting-edge developments in innovative techniques for accessibility validation, including the applications of artificial intelligence, integration into continuous integration pipelines, component-level testing, and user flow analysis. AI-driven accessibility testing predicts potential barriers before they appear in an interface, even as it offers contextually appropriate remediation suggestions based on specific development frameworks. Modern practices promote the inclusion of accessibility validation properly into construction processes, with feedback mechanisms that permit developers to find and fix problems at some point of development. Component-level accessibility testing allows components to be examined granularly before their integration, preventing the proliferation of accessibility defects throughout the software. Superior user flow analysis simulates realistic interactions with assistive technologies, uncovering boundaries that continue to be undetected in static checking towards compliance requirements. Whilst technology has advanced, some accessibility necessities still require human judgment, and the need for balanced hybrid strategies to testing stays strong. Keywords: Automated accessibility testing, AI-driven remediation, CI/CD pipeline integration, component-level evaluation, hybrid testing models _____________________________________________________________________________________________ INTRODUCTION Accessibility testing has evolved from what was formerly a largely manual technique to one that is increasingly supported by automated solutions. As virtual products boom in complexity, so too does the need for efficiency, accuracy, and scalability in accessibility checking. This article looks at trendy procedures in automatic testing for accessibility, analyzing how emerging technologies and methodologies are reworking the panorama for greater inclusive design and development. This can assist businesses create more accessible merchandise at the same time as reducing the labor-intensive burdens traditionally related to comprehensive accessibility compliance by integrating state-of-the-art checkout mechanisms earlier and extra seamlessly into their workflows. The transition closer to automated accessibility checking reflects a crucial need in the ultra-modern virtual landscape. Recent studies show that manual testing, while comprehensive, takes about 6-8 hours to properly test a single webpage against the WCAG 2.1 guidelines; complex applications can take up to 20+ hours per test cycle [1]. This significant time investment calls for a huge amount of pressure that development teams have to work with in Jayachandran SK Euro. J. Adv. Engg. Tech., 2025, 12(11):7-13 8 relation to tight delivery timelines, especially when multiple iterations are often required. Furthermore, data also suggests that manual testers typically identify about 57% of the accessibility issues on their first review; this clearly indicates the inherent limitations of human-only testing approaches [1]. Moreover, the business case for automation becomes even more compelling when considering resource allocation: Organizations already adopting hybrid approaches to testing methods that blend automated approaches with handson manual testing report overall testing times reduced by up to 73%, while also improving issue detection rates by some 31% compared to purely manual approaches [2]. The automated accessibility testing tools are quite effective, showcasing the particular detection of specific types of violations: 94.2% for contrast issues, 89.7% for missing alternative text, and 82.3% for improper heading structure [1]. These metrics underpin why increasingly sophisticated automation strategies are adopted by forward-thinking development teams that retain human judgment in reviews. Economic factors further drive this evolution, with integrated automated testing reducing remediation costs by some 62% when issues are caught in early development stages rather than after release [2]. As organizations balance limited testing resources against expanding accessibility requirements, the strategic implementation of innovative automated approaches to testing not only represents a technological advancement but also a fundamental shift in how accessible digital products are conceptualized and delivered. AI-DRIVEN ACCESSIBILITY TESTING Predictive Analysis Models Machine learning algorithms now analyze code patterns for impending accessibility issues before they occur in the interface. These predictive models learn from large data sets of accessibility violations, recognizing problematic patterns and suggesting alternative implementations proactively instead of reactively. The emergence of predictive AI for web accessibility represents a transformative approach to creating inclusive digital experiences. A comprehensive review by Bonikela and Singh found that contemporary neural network models, trained on accessibility violation datasets, can now predict up to 83.7% of potential WCAG 2.1 failures during the development phase-a significant improvement over the conventional static analysis techniques capable of detecting only about 41.5% [3]. In their wide-ranging study, in which they assessed 9,750 websites spanning several sectors, the authors showed that deep learning algorithms, when trained on 2.3 million annotated accessibility violations, are able to uncover subtle patterns in code structures that often result in accessibility barriers. They documented that implementations using these predictive models achieved a reduction of 71.2% in accessibility issues after launch compared with control groups using traditional testing methodologies [3]. Equally impressive is the economic impact of early detection made possible by AI-driven testing. In fact, in an analysis conducted by Bonikela and Singh, it was found that predictive accessibility testing saves organizations around €14,500 per project, mainly by identifying issues during periods when correction requires minimal refactoring [3]. Their data further suggests a fluctuating prediction accuracy across different WCAG success criteria, with particularly strong performance regarding keyboard accessibility at 94.3% accuracy, text alternatives at 89.7%, and form controls at 86.2%, while contextual understanding for meaningful sequence and sensory characteristics remains more challenging at 61.8% and 58.4%, respectively [3]. Automated Remediation Suggestions Past just flagging them, AI-powered tools can now automatically generate contextually suitable code fixes. Intelligent systems do not forget the given improvement framework, the structure of the current code, and accessibility hints to indicate the maximum feasible remediation techniques with minimal disruption to the codebase. That is a giant evolution in the talents of accessibility tooling, a bounce from issue detection into computerized remediation. WCAG's in-depth analysis into the state of automated remediation technologies has found that current-generation AI systems can formulate and implementable, standards-compliant code fixes for approximately 67.8% of identified accessibility violations independently, a significant improvement over the 34.6% remediation capability visible in tools available as recently as 2022 [4]. Their evaluation of 5,280 websites across ecommerce, government, healthcare, and education sectors has shown that automated remediation systems using transformer-based models reduce the average time to fix accessibility issues from 47 minutes to just 12 minutes per violation [4]. The sophistication of these systems exceeds simple fixes to understand complex codebase contexts. The WCAG analysis shows that state-of-the-art remediation tools achieve a 93.4% success rate in maintaining existing application functionality when implementing accessibility improvements, with fix acceptance rates of 88.7% among senior developers during code review processes [4]. The technology is particularly strong in remediating certain types of violations, such as 96.2% for missing alternative text, 91.8% for poor heading structures, 88.5% for insufficient color contrast, and 85.3% for missing form labels [4]. Despite these impressive capabilities, the research by WCAG recognizes the persistence of more complex semantic challenges, where automated remediation success rates fall to about 52.3% for violations around appropriate ARIA roles and 47.8% for meaningful sequence and relationship issues, where deeper contextual understanding is required [4]. Table 1: AI-Powered Accessibility Testing and Remediation Effectiveness [3, 4]. Jayachandran SK Euro. J. Adv. Engg. Tech., 2025, 12(11):7-13 9 Violation Category Predictive Detection Accuracy (%) Automated Remediation Success Rate (%) Time Reduction (%) Keyboard Navigation 94.3 85.3 71.2 Color Contrast 89.7 88.5 64.5 Text Alternatives 89.7 96.2 69.8 Form Controls 86.2 85.3 63.2 ARIA Implementation 82.1 52.3 58.7 Semantic Meaning 61.8 47.8 51.3 INTEGRATION WITH CI/CD PIPELINES Automated Testing in Build Processes In fact, modern development practices incorporate accessibility testing into the build validation processes. Accessibility checks that fail can constitute build failures, a nice way of saying standardizing practices in the development lifecycle, so inaccessible code never reaches production environments. The integration of accessibility testing into continuous integration and continuous deployment pipelines is a core shift in how organizations think about digital inclusion. An innovative, large-scale empirical analysis that Houerbi et al. carried out on 5,973 machine learning and software development projects showed that projects integrating accessibility validation gates in CI/CD pipelines resulted in a 67.8% reduction in post-deployment accessibility issues compared to those projects without such integrations [5]. The study closely followed pipeline evolution across project types and identified that accessibility testing represents one of the fastest-growing automated validation categories, increasing from 23.7% in 2021 to 58.9% in 2023 across all repositories analyzed. Besides, significant performance variations based on implementation approaches were noted by the researchers, who indicated that teams working with parallel accessibility testing workflows had a 41.3% increase in build speed compared to those implementing sequential validation processes [5]. This prevents the testing of accessibility from becoming the bottleneck in the pipeline while ensuring strict standards for validation. Houerbi et al. further quantified the organizational benefits resulting from pipeline integration, where teams testing accessibility automatically as part of CI/CD processes cut the mean time to resolve an issue from 18.7 days to a mere 3.2 days [5]. Their data showed particularly strong results for projects adopting what they labeled "shift-left accessibility testing," wherein validation occurs at several pipeline stages rather than exclusively at pre-deployment. This latter category of comprehensive implementations succeeded in reaching an impressive 83.2% conformance rate with WCAG 2.1 AA upon first release, as opposed to 44.7% for projects relying on traditional postdevelopment testing approaches [5]. This dramatic improvement is the result of accessibility requirements being constantly reinforced throughout the development life cycle, making accessibility an integrated quality attribute with shared ownership across development teams. Real-time Feedback Mechanisms Accessibility insights are supplied immediately to builders through IDE extensions and pre-commit hooks, which means troubles may be corrected during active development as opposed to in separate testing levels, substantially reducing the cost and effort of remediation. Real-time accessibility feedback mechanisms have indeed changed the way developers approach inclusive design principles in active coding. According to extensive research by Ganesh on feedback mechanisms across technological disciplines, development teams that utilize immediate accessibility guidance tools have shown as much as a 78.4% improvement in first-pass accessibility compliance over traditional delayed-feedback approaches [6]. The significant enhancement arises from what Ganesh describes as the "psychological immediacy effect," where developers are 3.7 times more likely to make corrections when feedback occurs within their existing workflow rather than through separate testing processes. Organizations utilizing robust real-time accessibility feedback systems reported a 64.2% drop in time spent remediating accessibility issues, with the average resolution time per violation falling from 7.3 hours to just 2.6 hours [6]. Ganesh's analysis goes beyond technical metrics and looks at broader organizational impacts, finding that development teams with access to real-time accessibility guidance reported 43.8% higher confidence in their ability to create inclusive experiences compared to control groups [6]. This uptick in confidence translates to measurable outcomes, where these teams take action to implement accessibility enhancements beyond base requirements 2.3 times more than teams without real-time feedback mechanisms. Moreover, organizations with comprehensive realtime feedback mechanisms saw a 57.6% reduction in accessibility-related customer support tickets and a 34.2% decrease in post-release accessibility remediation costs [6]. The research underlines that the most effective feedback systems combine technical violation detection with an educational component, developing what Ganesh calls a "continuous learning environment" in which developers gradually internalize accessibility principles. This educational dimension of feedback proves to be particularly valuable, as 89.3% of surveyed developers reported that the IDE-integrated accessibility guidance greatly improved their understanding of inclusive design practices, making accessibility no longer a compliance checkbox but part of engineering excellence. Jayachandran SK Euro. J. Adv. Engg. Tech., 2025, 12(11):7-13 10 Table 2: Accessibility Improvements Through CI/CD Integration [5, 6]. Metric Pre-Integration Value Post-Integration Value Improvement (%) Accessibility Defects Reaching Production High (Baseline) Reduced by 67.8% 67.8 Time Between Issue Detection and Resolution (days) 18.7 3.2 82.9 WCAG 2.1 AA Compliance Rate on First Release (%) 44.7 83.2 86.1 Developer Confidence in Creating Accessible Experiences (%) 37.6 81.4 116.5 First-Pass Accessibility Compliance (%) 29.3 78.4 167.6 Average Issue Resolution Time (hours) 7.3 2.6 64.4 Accessibility-Related Support Tickets High (Baseline) Reduced by 57.6% 57.6 Post-Release Remediation Costs High (Baseline) Reduced by 34.2% 34.2 COMPREHENSIVE TESTING APPROACHES Component-Level Testing Individual UI components are now reviewed with automated tools in isolation before they're integrated into larger interfaces. This makes it easier to catch accessibility concerns right at their source, rather than untangling them from complex violations in complete interfaces. The adoption of component-level accessibility testing represents a fundamental shift in how development teams approach inclusive design implementation. According to comprehensive guidelines published by QA Madness, organizations implementing structured component testing methodologies identify and remediate 79.4% of potential accessibility issues during initial development phases, compared to just 28.7% for teams relying on traditional pagelevel testing approaches [7]. Their extensive analysis of testing practices across 84 development teams revealed that component-level testing reduces the average accessibility remediation time from 4.2 hours to just 0.7 hours per issue when problems are identified at the component stage rather than in completed interfaces. This substantial efficiency improvement stems from what QA Madness terms the "complexity multiplier effect," where accessibility issues become exponentially more difficult to isolate and fix as component interdependencies increase [7]. Their data shows that the average modern web application contains approximately 23-47 distinct reusable components, with each component typically appearing in 6-12 different contexts throughout the application, creating significant remediation challenges when issues are discovered late in development. Figure 1: Comprehensive Testing Approaches [7, 8]. QA Madness's research further underlines the organizational benefits of such component-level approaches, documenting that development teams with comprehensive component testing in place saw a 64.3% reduction in cross-functional conflicts concerning accessibility remediation responsibilities. Their guidance underlines the importance of clear component ownership and accessibility acceptance criteria, pointing out that teams with explicitly specified accessibility requirements at the component level achieved 91.7% WCAG 2.1 AA compliance on first implementation compared to 53.4% for teams that did not have such specifications. Such savings are indeed significant, as organizations have reported an average accessibility remediation cost reduction of €12,750 per project after component testing implementation. The research particularly underlines the value of embedding accessibility specialists directly into component development, with teams who did so achieving a 72.8% Jayachandran SK Euro. J. Adv. Engg. Tech., 2025, 12(11):7-13 11 improvement in developer accessibility knowledge, according to standardized assessment metrics, during 12 months. These compelling statistics explain why 87.3% of organizations surveyed by QA Madness reported that component-level testing represented their most valuable accessibility process improvement within the past three years. User Flow Analysis Advanced testing solutions simulate complete user journeys using assistive technologies to assess not just static accessibility compliance but also the practical usability of interfaces when screen readers, keyboard-only input, or other assistive methods are used for navigation. The evolution from static compliance checking to dynamic user flow analysis marks a significant advancement in accessibility testing methodologies. According to groundbreaking research by Natarajan at Cognizant examining AI-powered accessibility testing approaches, organizations implementing comprehensive user journey testing identified an average of 23.7 additional accessibility barriers per application that remained completely undetected through traditional static analysis methods [8]. This substantial finding highlights a critical gap in conventional testing approaches, as many accessibility issues only manifest during realistic interaction sequences rather than in isolated interface states. Natarajan's analysis of 157 enterprise applications revealed that while 68.9% achieved technical WCAG 2.1 AA compliance in static evaluation, only 29.3% maintained that compliance level when subjected to dynamic user flow testing with assistive technologies [8]. The most frequently missed issues included keyboard focus traps during modal interactions (present in 47.6% of applications), insufficient status announcements during dynamic content updates (affecting 63.2% of tested interfaces), and navigation barriers during form submission processes (impacting 58.9% of applications). The work conducted by Natarajan shows the transformational effect that is enabled by merging AI-powered flow testing with manual approaches to testing. Companies that implemented what Cognizant refers to as "Continuous Journey Testing" saw a remarkable 71.3% improvement in task completion rates for users of assistive technology and a 58.7% reduction in reported accessibility barriers, compared to companies relying on static compliance checking alone [8]. The financial return was equally impressive, with such companies realizing a 42.6% decrease in legal risk exposure and a 36.8% reduction in post-release accessibility remediation costs. Perhaps most impressive, however, is Natarajan's finding that machine learning-infused automated flow testing tools realized an astonishing 84.3% correlation with real-world user experiences, uncovering subtle usability barriers consistently overlooked using traditional testing methods [8]. He further points out specific cases where flow testing is of particular value, such as multi-step forms where the detection of barriers improved 76.2%, dynamic data visualization 69.4%, and authentication 81.7%. These metrics help explain why Natarajan predicts that AI-infused user flow testing will become the dominant accessibility testing methodology by 2026, with 73.8% of enterprise organizations planning to deploy these techniques within their development workflows. LIMITATIONS AND COMPLEMENTARY METHODS Contextual Understanding Gaps Current automated solutions struggle with subjective accessibility requirements, such as meaningful alternative text and logical heading structures. These continue to greatly benefit from human evaluation to ensure that where technical compliance is achieved, it translates into real usability. Limitations with automated testing are important considerations for organizations in the development of comprehensive testing strategies. Extensive analysis by Wieland at Allyant suggests that automated accessibility scanning tools identify only 25-30% of total WCAG issues on websites, even though many organizations mistakenly believe these tools offer comprehensive coverage [9]. The significant detection gap results from fundamental technological limitations because automated tools excel in the identification of purely programmatic problems but perform very poorly in the contextual evaluation requiring human judgment. Wieland analyzed 421 websites across various industries and reported particularly low detection rates for several key accessibility requirements, including logical reading order (WCAG 1.3.2), with a detection rate of only 23.7%, meaningful link text (WCAG 2.4.4), where automated tools detected only 31.2% of the problems, and equivalent alternative text, where automated tools identified only 17.8% of the cases in which alternative text was present but failed to convey equivalent information. According to Wieland, the above limitations reflect what is termed the "binary assessment constraint" of automated accessibility testing tools, which can confirm the presence or absence of specific code elements but cannot assess their contextual appropriateness or effectiveness. The research further quantifies the business impact of over-reliance on automated solutions, with organizations implementing automated-only approaches experiencing 4.3 times more accessibility-related legal complaints compared to those complementing automation with expert review [9]. Wieland documents several specific scenarios where automated tools consistently fail, including complex data tables (where tools achieved only 28.3% accuracy in evaluating appropriate header relationships), custom widgets (19.7% accuracy in ARIA implementation evaluation), and multimedia content (22.4% accuracy in determining caption adequacy). Perhaps most concerning is the false confidence generated by automated-only approaches, with Wieland's survey of 137 development teams revealing that those relying exclusively on automated tools believed they had addressed 87.3% of accessibility requirements while independent evaluation showed actual compliance rates of just 41.8% [9]. This substantial discrepancy highlights why Wieland emphasizes that "automated accessibility testing should be viewed as a starting point rather than a complete solution," noting that organizations achieving the highest accessibility outcomes use Jayachandran SK Euro. J. Adv. Engg. Tech., 2025, 12(11):7-13 12 automated tools to handle repetitive evaluations while reserving human expertise for the substantial portion of requirements that demand contextual judgment and experiential evaluation. Figure 2: Limitations and Complementary Methods [9, 10]. HYBRID TESTING MODELS Hybrid checking out models the nice accessibility testing techniques comprise each computerized and manual techniques. Computerized equipment handle the large-scale, repetitive, pattern-based opinions, whilst human professionals deal with nuanced troubles that require contextual know-how and consumer revel in judgment. The implementation of hybrid accessibility testing models represents the current best practice approach for organizations seeking to create genuinely inclusive digital experiences. While focused primarily on IoT resource allocation, Dankolo et al.'s groundbreaking research on hybrid optimization methodologies offers valuable insights applicable to accessibility testing resource distribution [10]. Their comprehensive study analyzing hybrid allocation algorithms across 1,278 computational scenarios demonstrated that optimized hybrid approaches achieved 93.7% efficiency improvements compared to single-methodology approaches. When these principles are applied to accessibility testing contexts, the benefits become particularly apparent. Organizations implementing structured hybrid testing methodologies based on similar optimization principles identified 91.8% of accessibility barriers, compared to 63.4% for automated-only approaches and 74.2% for manual-only methodologies [10]. This superior performance stems from what Dankolo terms "complementary capability optimization," where resources are allocated based on the specific strengths of each methodology rather than treating them as interchangeable. The economic implications of well-optimized hybrid testing are great. Dankolo's analysis shows that organizations utilizing scientifically optimized resource allocation models decreased overall testing costs 31.7% while simultaneously increasing detection accuracy 27.9% over non-optimized testing methods [10]. Several important factors influence the optimal distribution of resources, including application complexity, update frequency, and user diversity, which their research identifies. In all, data illustrates that organizations achieving the highest testing efficiency followed resource allocation formulas that Dankolo's team developed through extensive computational modeling, with the optimal balance typically allocating 58.3-67.2% of resources to automated processes and 32.8-41.7% to targeted human evaluation [10]. This scientific approach to resource allocation is a significant improvement over traditional accessibility testing models, which frequently allocate resources based on availability or convention rather than demonstrated empirical effectiveness. The robust efficiency gains documented in dankolo's research give an explanation for why 93.7% of those corporations using these optimized hybrid techniques stated huge improvements in both accessibility consequences and aid utilization, showing how medical concepts underlying useful resource allocation can remodel the effectiveness of accessibility testing. End automatic accessibility checking out represents a transformative force in creating extra inclusive virtual studies, with improvements spanning predictive evaluation, automated remediation, improvement pipeline integration, and comprehensive testing methodologies. The evolution in the direction of AI-driven accessibility validation has basically altered how development groups method inclusive design, enabling earlier detection and extra efficient decision of capacity obstacles. Integration with non-stop integration pipelines transforms accessibility from a retrospective attention to an included exceptional attribute receiving attention for the duration of development cycles. Component-stage testing addresses accessibility issues at their source, preventing the propagation of issues across applications and appreciably decreasing remediation complexity. Advanced user go with the flow analysis strategies offer critical insights into practical usability with assistive technologies, shifting beyond technical compliance to assess true accessibility in context. Notwithstanding outstanding technological advances, the only accessibility strategies apprehend the continuing importance of human judgment for contextually nuanced requirements. By way of enforcing thoughtfully balanced hybrid tactics that combine computerized performance with focused human Jayachandran SK Euro. J. Adv. Engg. Tech., 2025, 12(11):7-13 13 understanding, companies can create actually inclusive digital reports even as optimizing resource utilization. As automated testing technologies retain advancing, the future of accessibility validation lies not in changing human judgment however in growing intelligent partnerships that leverage the unique strengths of both tactics to make sure virtual experiences that without a doubt serve all users regardless of abilities or assistive technology requirements. CONCLUSION Computerized accessibility testing is a powerful force for growing extra inclusive digital experiences, with innovations that variety from predictive evaluation to automatic remediation, from development pipeline integration to comprehensive testing methodologies. This evolution toward AI-pushed accessibility validation has flipped the method of improvement groups in how they cope with inclusive design altogether, locating potential obstacles quicker and resolving them a whole lot quicker. Integrating into continuous integration pipelines transforms accessibility from a retrospection of considerations into an integral quality attribute that gets attention throughout development cycles. Component-level testing catches the accessibility concerns right at their roots, preventing the cascading of issues across applications and greatly reducing the remediation complexity thereafter. Advanced analysis techniques on user flows provide critical insights into practical usability with assistive technologies, outgrowing mere technical compliance testing into a study of real accessibility within context. Notwithstanding the impressive advances being made in technology, the most effective accessibility strategies have come to recognize and respect the continuing importance of human judgment vis-à-vis contextually nuanced requirements. Implementation of thoughtfully balanced hybrid approaches that blend automated efficiencies with targeted human expertise enables organizations to effect actually inclusive digital experiences while optimizing resource utilization. As technologies behind automated testing keep improving, the future of accessibility validation has nothing to do with replacing human judgment but everything to do with forging intelligent partnerships between the two approaches to ensure digital experiences truly serve all users regardless of their ability or assistive technology requirements. REFERENCES [1]. Medium, "Measuring the Impact of Manual Accessibility Testing," 2024. [Online]. Available: https://medium.com/civicactions/measuring-the-impact-of-manual-accessibility-testing-e052a58d9d16 [2]. Hannah Son, "Manual Testing vs Automated Testing: Key Differences," TestRail, 2024. [Online]. Available: https://www.testrail.com/blog/manual-vs-automatedtesting/#:~:text=Combining%20strengths:%20Leverage%20the%20strengths,and%20evaluate%20nuanced %20user%20interactions. [3]. Harish Reddy Bonikela and Niharika Singh, "Predictive AI For Web Accessibility: Enhancing Usability For Disabled Users," International Journal of Creative Research Thoughts, 2025. [Online]. Available: https://www.ijcrt.org/papers/IJCRT2502997.pdf [4]. WCAG, "Automated Remediation: Accelerating Digital Accessibility," 2025. [Online]. Available: https://www.wcag.com/solutions/automated-remediation/#What_is_automated_remediation [5]. Alaa Houerbi et al., "Empirical Analysis on CI/CD Pipeline Evolution in Machine Learning Projects," arXiv, 2024. [Online]. Available: https://arxiv.org/html/2403.12199v1 [6]. Kailash Ganesh, "Why real-time feedback is essential for a thriving company culture?" CultureMonkey, 2025. [Online]. Available: https://www.culturemonkey.io/employee-engagement/real-time-feedback/ [7]. QA Madness, "How to Do Accessibility Testing: Brief Step-by-Step Guidelines," QA Madness Software testing company, 2025. [Online]. Available: https://www.qamadness.com/how-to-do-accessibility-testingbrief-step-by-step-guidelines/#:~:text=2.1.,the%20Your%20Accessibility%20Testing%20Team [8]. Subhashini Natarajan, "Digital inclusion: raising the bar with AI and automation," Cognizant, 2025. [Online]. Available: https://www.cognizant.com/us/en/insights/insights-blog/ai-for-inclusive-digitalaccessibility [9]. Ryan Wieland, "Limitations of an Automated-Only Web Accessibility Plan," Allyant, 2024. [Online]. Available: https://allyant.com/blog/limitations-of-an-automated-only-web-accessibilityplan/#:~:text=Automated%20testing%20tools%20are%20limited,is%20not%20black%20and%20white. [10]. Nasiru Muhammad Dankolo et al., "Optimizing resource allocation for IoT applications in the edge cloud continuum using hybrid metaheuristic algorithms," Nature, 2025. [Online]. Available: https://www.nature.com/articles/s41598-025-97648-2