What Teams Say After Working With Us
Genuine feedback from Malaysian software teams on what the engagements looked like in practice.
Back to Home6+
Years in business
120+
Engagements completed
4.7
Average satisfaction
94%
Clients returning
Feedback From Real Engagements
These reviews reflect actual project experiences from clients across Kuala Lumpur, Petaling Jaya, and the wider Klang Valley area.
Siti Khadijah
CTO · Fintech startup, KL
"We brought in Neuraplex after a major release caused a number of regressions in our payment flow. The test automation framework they set up has genuinely changed how we approach releases — we now have meaningful coverage in areas we'd been avoiding because they were too complex to test manually."
March 2026 · AI Test Automation
Farid Hadzir
Lead Dev · E-commerce platform, PJ
"Visual regression testing was something we kept saying we'd get around to. The Neuraplex team made it a lot less daunting than I expected — the baseline setup took most of one week, and by the second week we were already catching visual inconsistencies in a redesigned checkout section that would have been noticed by customers first."
February 2026 · Visual Regression
Lim Wei Yang
Engineering Manager · SaaS, Cyberjaya
"We engaged Neuraplex for the defect prediction work. The model they built surfaces risk indicators before each sprint review — it's not always right, but it's directionally useful and has shifted how our planning conversations go. The dashboard is readable even for people who aren't engineers, which I appreciated."
March 2026 · Defect Prediction
Nurul Rahim
QA Lead · Enterprise software, Shah Alam
"The thing I remember most clearly about this engagement is how thoroughly everything was documented. When the Neuraplex team handed over the test framework, I could actually follow what had been built and why — which made it far easier to maintain and extend on our own."
February 2026 · AI Test Automation
Azlan Tahir
Product Manager · Mobile app, Subang
"For our mobile app we were getting complaints about UI inconsistencies across device sizes. The visual regression setup Neuraplex delivered now runs automatically after each build. We haven't had a customer-reported visual bug in the three months since it went live. That's the clearest outcome I can describe."
January 2026 · Visual Regression
Chan Jin Hao
Director of Engineering · Marketplace, KL
"We used the defect prediction service ahead of a significant platform migration. The risk dashboard flagged two modules as high-concern that we hadn't flagged ourselves — and both of those did produce defects in early testing. It's not a crystal ball, but it added a useful layer to our pre-release process."
March 2026 · Defect Prediction
Projects in Detail
Three engagements described in more depth — what the situation was, how we approached it, and what changed.
Fintech Platform CI Integration
AI Test Automation · 4 weeks
CHALLENGE
A fintech team in Kuala Lumpur was releasing every two weeks but spending three to four days of each cycle on manual regression testing. Coverage was inconsistent and error-prone, particularly in authentication and payment flows.
OUTCOME
AI-driven test automation with CI integration reduced manual regression time from 3–4 days to under 4 hours per release cycle. Test coverage of payment and authentication modules increased from 41% to 87%.
87%
coverage
4h
regression cycle
E-Commerce UI Consistency
Visual Regression · 2.5 weeks
CHALLENGE
An e-commerce platform in Petaling Jaya was receiving customer complaints about inconsistent button styles, broken layout on Samsung devices, and occasional text truncation on product cards — introduced during a design system overhaul.
OUTCOME
Visual regression suite now runs on every build, catching 15+ visual inconsistencies in the first month. No customer-reported visual bugs in the four months following deployment.
15+
bugs caught/month
0
customer UI reports
SaaS Pre-Release Risk Model
Defect Prediction · 5 weeks
CHALLENGE
A B2B SaaS team in Cyberjaya was experiencing unpredictable release quality — some sprints went to production cleanly, others required multiple hotfixes. They had good defect data but no way to use it predictively.
OUTCOME
Defect prediction model now scores each release candidate before testing begins. Model accuracy reached 73% on first validation. Post-release hotfix rate dropped 38% in the two quarters following deployment.
73%
model accuracy
38%
fewer hotfixes
Get in Touch
Office
Damansara Utama,
Petaling Jaya, Selangor
Hours
Mon–Fri: 9am–6pm
Sat: 9am–1pm (MYT)
Your Team's Experience Could Be Next
If any of the above situations sound familiar, a conversation with us is probably worth the time.
Start a Conversation