I track emerging shifts, then build to understand them. Right now, that means Generative AI.
Track emerging shifts early, while they are still ambiguous.
Where AI changes how products are made, and what products can do.
Build tools and prototypes with them, to learn what they really change.
What those shifts do to teams, organisations, and digital systems.
Understand their second and third order effects on teams and organisations.
Which trade-offs, unintended consequences, and strategic bets matter.
The patterns I keep returning to
Pick a pattern below to see the writing and projects behind it.
Recurring threads
Tracing this pattern
Applied AI in practice
AI work here starts from real constraints: confidential data, uneven adoption, messy files, cost, latency, and the cognitive and intent debt these tools can create. The pattern is turning model capability into something usable, inspectable, and safe enough for actual work.
In the Journey
- Applied Innovation Lead, Corporate Strategy & AIJun 2023 — now · Ufinity, Singapore
Writing
- One bug, eight fixes, and the cost of cognitive debtAI can accelerate implementation and repairs, but it does not replace the deliberate thinking needed to catch requirements that were never stress-tested.
- How you talk to AI is itself a design decisionSpec-driven prompting and iterative prompting are both design choices. The useful question is where the quality judgment lives: in a contract you can define upfront, or in feedback you only get once something exists.
- AI is a universal translator, but it still amplifies your intentAI compresses the distance between intent and working software. That makes small tools worth building, but it also makes unclear intent show up faster and louder.
Projects and tools
- Now In SingaporeA curated pulse-check on what is happening across Singapore right now.
- PromptPalAn internal Generative AI product for confidential, organisation-wide LLM access.
- PromptPal TranscribeA focused transcription tool that turns spoken work into usable written material.
The ideas I sit with
All writing →- One bug, eight fixes, and the cost of cognitive debtAI can accelerate implementation and repairs, but it does not replace the deliberate thinking needed to catch requirements that were never stress-tested.
- How you talk to AI is itself a design decisionSpec-driven prompting and iterative prompting are both design choices. The useful question is where the quality judgment lives: in a contract you can define upfront, or in feedback you only get once something exists.
- AI is a universal translator, but it still amplifies your intentAI compresses the distance between intent and working software. That makes small tools worth building, but it also makes unclear intent show up faster and louder.