If people cannot read the system, the system is too expensive.
I have seen teams spend too much energy decoding the work before they can do the work. A cleaner surface helps only when the important parts become easier to see.
About / Operating pattern
I work on the systems that make judgment easier to see.
My career has moved from making digital experiences usable, to helping teams and products make better decisions, to exploring how AI systems can stay useful without making people smaller.
The through-line is not a job title. The recurring pattern is making the hidden load visible enough for people to work with.
Career evolution
From usable screens to clearer decisions.
01
I started with hands-on craft: web, mobile, brand, learning products, campaign systems, and the practical discipline of making digital work clear enough to use.
02
That craft expanded into systems: product foundations, regional consistency, design governance, training, service localization, and operating structures that help teams work with less ambiguity.
03
At Indeed, the work sharpened around employer workflows, recruiting automation, mobile strategy, product-learning programs, research synthesis, and decision-load reduction.
04
Now the same pattern moves into AI: personal operating systems, agent coordination, approval gates, local-first private work, evidence-based product intelligence, and human-controlled automation.
Operating beliefs
Technology should not make people smaller.
This belief shows up before AI: in hiring products, mobile strategy, design systems, training, facilitation, and team capability work. SEEK was an early version of the pattern: regional marketplace systems, employer and candidate surfaces, staging and release quality, and cross-market product delivery all had to move together.
Automation should show its hands.
When a system acts for someone, I want to know what it did, why it did it, where it stops, and how a person can take control again.
Evidence is not direction yet.
Research, metrics, feedback, and product signals can still leave a team stuck. The useful moment is when the evidence becomes choices people can argue with.
Good design should leave people more capable.
In the strongest evidence here, that has meant an interface, a workshop, a critique habit, a training system, or a way for the room to decide without pretending the work is simple.
How I work
I look for the hidden load.
Start by asking what decision the product or team is avoiding.
Turn scattered evidence into choices people can compare.
Make automation boundaries visible before asking anyone to trust the system.
Use workshops, advisor models, design systems, and rituals when the room needs better handles.
Keep craft visible, but do not ask craft alone to carry the whole story.
Career arc
The AI chapter is new. The pattern is not.
That is why this portfolio keeps team capability, facilitation, design systems, and training visible. They are where much of the pattern started.
Early career signal
Apple
High standards, service design, localization, training contexts.
Where the service and training standards came from: quality, localization, and human capability under real-world constraints.
Hands-on product craft
EF / Englishtown
Digital learning products, mobile/iPad experiences, CMS, and visual systems.
Early hands-on craft, shown as archive context rather than the main story.
Regional product/platform complexity
SEEK Asia
Regional marketplace product/design systems, cross-market product delivery, and design governance.
Early systems work across jobsDB / JobStreet hiring-marketplace surfaces: employer tools, candidate flows, responsive communications, company-profile experiences, staging/release coordination, and product delivery alignment across regional contexts.
Senior design leadership
Indeed
Hiring workflows, product strategy, mentoring, facilitation, and design maturity.
The main source of flagship cases: hiring workflows, product strategy, mentoring, facilitation, research synthesis, and design maturity.
Current direction
Habitat / KIREI / JAVIS / AI-native systems
Personal AI environment, human-centered automation, product intelligence, job intelligence.
Shows what I am testing now, without pretending active explorations are validated products.