Doubled delivery throughput
Consolidated fragmented codebases onto a shared platform and automated the delivery pipeline, cutting the engineering footprint required per release from ~80 to 25–30 engineers. The released capacity funded new product bets rather than headcount reduction.
- 2×delivery throughput
- ~80 → 25–30engineers per release
AI-assisted engineering, org-wide
Drove adoption of Cursor, MCP servers, autonomous bug agents and AI code review across the whole organisation — and authored the engineering standards governing AI usage, review and code security.
- %of pull requests AI-assisted
- → daysmedian cycle time
Web performance, rebuilt end-to-end
Took revenue-critical pages from failing to fast: sub-1 second P50 and sub-2 second P95 real-user load times, with Core Web Vitals passing on the majority of them.
- 28 → 80+Lighthouse score
- +%mobile booking conversion
- −%bounce rate
Sub-1MB Progressive Web Apps
Shipped PWAs under 1 MB with complete offline functionality, enabling lightweight distribution to partner channels and bandwidth-constrained markets.
- <1 MBapp payload
- 100%offline capable
Localisation & payments platform
Delivered a localisation and payments platform supporting multi-currency checkout across Southeast Asian markets, from a single shared codebase.
- 16languages
- markets supported
Built the organisation
Established the interview loop, career ladder and performance calibration process now used across the engineering function — and hired the managers who run it.
- engineers hired
- engineering managers hired
- % → %voluntary attrition
Executive partnership
Own quarterly roadmap negotiation and annual capacity planning with Product, Design, Data and Business leadership, reporting technology strategy and delivery health to .