Insights
Field notes on putting AI to work inside a business, and on software quality up close.
When AI builds it in two hours but verification still takes three days
Once AI speeds up development, spec sign-off, testing and release do not automatically speed up with it. Using a real case from a booking system SaaS, we describe how we adjusted the SDLC to cut full release verification from two or three days to under two hours, and how the SQA role changed along the way.

AI agents are everywhere. How should a small business choose one?We were recently asked to run a full post-deployment review of an AI agent, and found it had quietly caused five times too much raw material to be ordered from customer order emails, and nearly created a liquidity risk on the finance side. The problem with agents is rarely that they are not smart enough. It is that nobody asked the right questions.
Three patterns we see once AI coding reaches the teamAfter working with more than ten development teams over the past six months, we keep seeing the same thing: individuals getting faster with AI does not mean the team is ready to get faster together.
Why we rebuilt SQA.TWThe SQA.TW redesign is not us leaving SQA. It lets the site show more fully what we have always done, which is understand how a business operates and find where digitalisation and AI can genuinely help.
Two things we learned from a North American e-commerce projectBuilding a decision platform last year for cross-ocean replenishment and fulfilment risk across two North American warehouses, we confirmed again that the hard part is not plugging AI into a system. It is understanding the timing, constraints and consequences of every decision.
Qaptly Desktop now supports CodexFrom v0.17.5, the AI Assistant in Qaptly Desktop officially supports Codex. If you already have a ChatGPT subscription, you can now have the AI Assistant use Codex and save on costs in day-to-day test design and automation work.
Once AI joins the workflow, the things that matter most are the easiest to underestimateSince Lunar New Year we have mostly been helping a few startups and in-house development teams at larger companies adopt agentic development. We have seen a lot of different approaches in that time and come away with some very direct impressions, so here they are.
From point tools to closed-loop processes: what is real and what isn'tA look at how AI is used across the software development lifecycle in 2026, the shift from widespread point tools to closed-loop collaboration across the process, and the real and false boom in QA automation.
It turns out a testing tool can be for fun tooWhy we decided to let users skip the Accessibility check, the engineering and product thinking behind that decision, and how Qaptly Auto for Android keeps both flexibility and clear limits between formal testing and less typical uses.
When software testing meets the real worldOnce a system starts interacting with the real world, verifying screens and APIs is no longer enough.
This article explains why Qaptly opened up UPP (Unified Pilot Protocol),
and how combining it with natural language lets IoT tests verify whether things actually happened.
[A hidden SQA indicator] Structured error codesError codes are not just an error-handling tool but an important indicator of software quality maturity. An SQA view of how error codes affect diagnosis speed, team collaboration and user experience.
Qaptly is here!Qaptly is an important milestone on our way to building a smart QA tool. v0.13.0 should save SQA engineers a good chunk of their lives again. Leave the execution to us, free up your hands, and go do something more meaningful.
Is using AI the line between development teams that survive and those that don't?AI is changing how development teams pace and coordinate their work. Teams that integrate AI well usually deliver more, and more stable, output in less time, rather than buying efficiency with overtime. The real gap is not the technology itself but a change in mindset and habits. People who know how to work with AI will redefine what development efficiency means.
Why we don't recommend building or outsourcing your own e-commerce siteBuilding your own e-commerce site looks cheaper than handing a cut of every sale to a marketplace. This article explains why we still advise against it.
A simple way to check the quality of a project or productIf you are not a software developer, it is hard to judge the quality of a project or product. This time we skip the technical detail and look at simple methods anyone outside engineering can use to assess their own project or product.
Qaptly now automates desktop apps on macOS 10.15+ and Windows 10/11Qaptly already had macOS and Windows versions, but they only automated website testing in Chrome/Edge on those platforms, plus Android/iOS testing. The latest version, 0.9.23, adds native desktop app support, and test cases can still be fully automated from natural-language descriptions.
Testability is the core of software qualitySoftware comes in many forms and is built in many ways, and SQA methodologies keep multiplying. But how you test is only technique; what matters is the underlying principle. If we had to pick the single most important principle, it would be testability. Unfortunately, traditional SQA rarely gets a say in it, and developers seldom consider testability early in system design.
Should information security be a required course?Any logic that looks complete and self-consistent has to be rethought for information integrity the moment a necessary third party joins in
What a system crash teaches us about stress testing mistakesHow we think about a system's processing and load capacity, using Taiwan's receipt lottery draw as the example.
How much testing makes sense? How SQA TW works and what our services look likeA quick guide to how SQA TW works day to day, how we price our work and how we use tools to support verification, for anyone thinking about working with us.
What inconsistent UI behaviour tells us about software quality managementA small UI consistency issue opens a look at what software quality management is really about. When similar input fields respond on different triggers, it is not just an interface issue. It points to deeper questions about code quality and design thinking.
Why does testing keep missing the problems?A look at why software testing keeps failing to catch problems, covering issue types, testing blind spots and how to improve. From the technical side to process, it helps teams build a more complete quality assurance system and find more defects.
You paid an outsourcing team a lot for your system. How do you avoid paying again for changes?Risk management for outsourced software development, from requirements analysis and system design to acceptance testing, with a practical approach to quality verification that helps companies avoid extra costs later.
SQA engineer or test automation engineer, which should you hire?How the roles of SQA engineer and test automation engineer differ and are changing in the age of AI. The two roles have traditionally had different responsibilities, but AI is changing that and making us
rethink how we choose testing talent.