A systems view of AI readiness
Before moving to the UK, I worked for Toyota China, where I practised Just-in-Time, kaizen, and problem visualisation. That experience shaped the way I look at business problems. A process is not just a workflow. It is a signal path. If the signal is broken, delayed, hidden, or ignored, the whole system begins to produce waste.
After moving to the UK, I worked with different businesses and saw a different pattern. Many organisations focused heavily on the balance sheet, but were slower to invest in technology, equipment, process design, and worker autonomy. Kaizen sounds attractive in theory, but it does not work well when the people closest to the problem do not have enough power to change the process.
Why AI can amplify hidden failure
AI does not simply automate work. It amplifies the system it is placed inside. If the workflow is clear, the data is reliable, and the decision path can be traced, AI can help.
But if the organisation already has hidden problems — poor data, disconnected teams, unclear ownership, manual workarounds, dashboard contradictions, or decisions no one can explain — AI can make those failures faster, larger, and harder to challenge.
What failure is the current process compensating for?
What Amazon taught me
For more than 15 years, I have traded on Amazon. At first, marketplace selling looked like a business of products, logistics, pricing, and marketing. Over time, it became something much more technical: a constant adaptation to platform rules, catalogue systems, pricing engines, policy changes, account-health controls, automation, and invisible decision layers.
Amazon is one of the clearest examples of platform-scale automation moving faster than accountability. As a small seller, I have witnessed how catalogue data, support workflows, pricing systems, compliance rules, and seller-facing dashboards can fail to line up. A seller may see one reality, support may quote another, and the automated system may act on a third.
What Y-Trace does
Through Y-Trace and SellerTrace, I investigate platform and marketplace failures that are difficult to see from one dashboard alone. I use AI-assisted analysis to organise messy evidence, trace contradictions, identify missing signals, and turn operational confusion into structured audit cases.
My focus is not simply whether a business can deploy AI. My work sits earlier: before automation, before model deployment, before a broken process is made faster by machine logic.
Trace the decision path
Which data, rule, dashboard, or system state produced the visible outcome?
Trace the failure path
Where did the signal break, disappear, loop, or contradict another source?
Trace the consequence path
Who carries the cost when the system cannot explain its own decision?
Trace before automation
Do not ask AI to scale a workflow until the hidden failure has been made visible.