Enterprises across Singapore and the wider Asia-Pacific (APAC) region are facing mounting losses from poor software quality, with Singaporean firms alone reporting annual costs of around USD 500,000 (S$600,000) due to fragile systems and technical debt. The issue stems largely from rushed software releases, untested code, and reliance on legacy infrastructure, which erodes customer trust and slows growth.
According to Damien Wong, Senior Vice President at Tricentis, over a third of global teams cite unresolved technical debt as the biggest barrier to software quality. Legacy systems, particularly monolithic applications and outdated infrastructure, further complicate integration with modern workflows. Forrester forecasts that software and IT services will account for 66% of global tech spending in 2025, much of it dedicated to modernisation.
Key strategies for enterprises include:
- Test gap analysis to identify untested high-risk zones in new code.
- AI-driven, model-based test automation, which reduces manual effort and adapts to system changes.
- Continuous quality engineering, embedding resilience early in development to catch defects before they escalate.
- Performance testing, treating latency or scalability issues as business risks since “slow is the new downtime.”
AI is also playing a pivotal role. Wong highlighted case studies such as Worldpay, which achieved a 90% reduction in maintenance effort by automating testing with Tricentis Tosca. AI-powered tools enable smoother legacy integration, prevent future debt through model-based testing, and democratise quality by allowing non-technical stakeholders to validate requirements. However, AI must be used with human oversight to avoid risks such as hallucinated outputs or misaligned results.
Wong stresses that CIOs should treat software quality as a business value driver, not a cost centre. Mature DevOps teams that embed AI in testing report being 30% more effective, cutting production errors, accelerating releases, and ultimately boosting revenue and customer loyalty.
Source:
