Urgent.News

What's breaking now, across hundreds of outlets.

Tech

Quality Debt Is the New Technical Debt

For years, we talked about technical debt as something that quietly piles up while teams ship fast and skip cleanup. You take a shortcut now, but you pay for it later, usually at the worst possible time. A newer version of that problem is showing up in software delivery, driven by AI coding agents and […]

Quality Debt Is the New Technical Debt

For years, discussions surrounding technical debt have centered on the consequences of taking shortcuts during software development, often leading to issues surfacing at the most inopportune moments. The emergence of a similar predicament in the realm of software delivery is being fueled by advancements in AI coding agents and the discovery of security vulnerabilities through AI.

Unfortunately, a notable trend has arisen where the focus on speed has overshadowed the emphasis on quality, giving rise to what is now termed as "quality debt." Companies are no longer grappling with the challenge of shipping software; instead, AI coding assistants have taken over that aspect. Development cycles have accelerated, with code changes now being released on a weekly, if not daily, basis.

However, traditional testing methods struggle to keep pace with this rapid pace. Quality assurance teams, still operating on Agile or DevOps cycles designed for a slower era, find the gap between the speed of code deployment and the verification process widening rather than narrowing. This conundrum is captured by the concept of quality debt.

While it is possible to deploy software through a CI/CD pipeline within an afternoon, there remains a significant lack of confidence in the code's improved state compared to its previous version.

To address this issue, UiPath has introduced a solution called Test Cloud, which was showcased in the Tech Field Day Showcase. The maturity model presented by UiPath offers a framework to gauge the current state of testing maturity within organizations. At its most basic level, manual testing involves individuals clicking through screens, hoping to remember what to check.

Moving up, scripted UI automation employs a virtual keyboard and mouse, often lacking comprehensive API and mobile coverage, and tends to be fragile and prone to frequent failures. The third stage involves testing that is genuinely embedded within the software development lifecycle, where testing is shifted "left" and integrated throughout the development process. At the pinnacle is continuous testing, where AI plays a role in adjusting activities in real-time.

UiPath's approach distinguishes between robots and agents. Robots execute test cases consistently and predictably, making them ideal for stable, repeatable regression testing. On the other hand, agents employ AI to reason through each step and adapt to changes in the workflow, which makes them more suitable for initial testing phases or scenarios that are ambiguous enough that a fixed script would fail upon encountering even minor changes.

A key takeaway from the showcase was UiPath's recommendation that organizations use robots for stable tasks and agents for less predictable ones. The concept of using AI as an augmentation rather than a replacement for deterministic automation is emphasized, as AI can enhance efficiency but should not take the place of essential, reliable testing processes.

Rewritten by urgent.news; may contain errors — check the original.

Sources: devops-hd5ff.sevalla.app · devops.com

This story

This is one outlet's version. Read the fullest account.

Every outlet that reported this, and when →

Read the original at devops-hd5ff.sevalla.app →

Report a problem with this story

More in Tech

More from Thursday 8 October →