Plexure · Software Engineer
Auckland, New Zealand · Dec 2022 - Present700M+ consumers83M consumers migrated24h → 3h processing174k+ users recovered
- Led a cross-functional team of 7 engineers from design through production rollout of a modernised loyalty points expiry service using Azure Functions, Cosmos DB, and Azure Container Apps, reducing a critical monthly processing workload from 24 hours to 3 hours for more than 5 million consumers.
- Owned technical design and implementation of the next phase of McDonald's loyalty API platform, re-engineering legacy capabilities into simpler domain services and improving response times by 3x to 5x while significantly reducing infrastructure cost.
- Drove end-to-end architecture for a loyalty reward control system across a distributed microservices architecture using C#/.NET, Cosmos DB, and ASP.NET Razor: designed market-level and consumer-level activation limits with cooldown logic to give marketers precise control over reward campaigns; led cross-team design reviews and authored decision logs to align on domain ownership.
- Optimised large-scale background processing from 8 hours to 10 minutes, significantly increasing throughput and reducing end-to-end system processing time.
- Co-led the design and execution of a complex migration of 83.3 million consumers to a single points-balance model, identifying delivery risks early and defining safeguards to protect data integrity during rollout.
- Built an internal remediation tool that restored loyalty points for 174k+ affected consumers, establishing a reusable recovery capability for future production incidents and improving incident resolution speed.
- Designed and implemented alerts-as-code using Terraform, New Relic, Azure Monitor, PagerDuty, Raygun, and Slack to standardise production visibility and improve incident response for multiple services.
- Drove shift-left engineering practices, including unit, component, and early integration testing, to improve release confidence and catch issues earlier in the delivery lifecycle.
- Led urgent production fixes for duplicate loyalty point allocation, coordinating release planning, QA validation, and post-deploy monitoring to quickly stop customer impact and reduce revenue risk.