Every stop added a layer.

Research discipline first (aviation, 2017)

I started as a commercial analyst covering the Asia-Pacific business aviation sector — industry reports, market analysis, structured databases. An unusual first stop, but it taught me the habit that underpins everything since: if the data isn't organised and the question isn't framed, the analysis is worthless.

Measurement rigour (adtech, 2018)

At Amnet (Dentsu Aegis Network) I learned how performance data is actually produced: tag implementation, data layers, marketing pixels, floodlight tags. When you've built the tracking yourself, you know exactly what a number can and can't tell you. Applying that rigour to a campaign cut an NGO client's cost-per-acquisition by 40%.

Healthcare domain and growth analytics (FindDoc, 2019–2023)

Four years at a Hong Kong healthcare platform running growth analytics for pharmaceutical clients — 20+ campaigns including Dupixent for Sanofi and PCV-13 for Pfizer. Weekly optimisation cycles in Looker Studio, budget advisory, analytics infrastructure built with engineers. This is also where I started automating my own job: Python scripts assisted by LLMs, scheduled on Google Cloud, cut reporting hours by 30%. Healthcare stopped being a client vertical and became my domain.

Evidence methods (MSc, 2024–2025)

Marketing analytics tells you what is happening; it rarely tells you why adoption fails. I went to the University of Groningen for an MSc in Business Administration (Health) to close that gap — health systems, health economics, operations — and graduated with Distinction. My thesis synthesised the evidence on why digital health interventions succeed or fail across contexts, and my modelling project compared vaccination strategies under uncertainty. Alongside the degree I worked part-time for Bowtie, Hong Kong's first virtual insurer, researching European insurtech strategy.

Regulated-industry rigour (compliance, 2026)

Currently at Maples Group, where I've pointed the same toolkit at compliance operations: Excel screening tools that speed up sanction checks, and Power Automate workflows that took the manual tracking out of case management. Different domain, same result — the team moves faster.

What this adds up to

Most analysts go deep on one stack in one industry. I've done something different: proven, five times over, that I can land in an unfamiliar domain and be useful fast — because the underlying craft transfers. If your team's data is fragmented, your reporting is manual, or your decisions need evidence behind them, that's the job I've been doing for eight years.