Moravy Oum

Moravy Oum · Interactive résumé

Don’t take my word for it. Ask.

I’m a software engineer at Xero. In eight months I’ve taken a new service from empty repo to production, owned incidents end-to-end, and helped drive production alerts from 258 a month to around 54, with AI as a force-multiplier, not a crutch. This page knows the details. Ask it anything.

Pick a question. Every answer is pre-written, no AI guessing

Recruiters & hiring
For engineers

Selected work, 8 months at Xero

  1. 01

    Ownership

    Took a new service from empty repo to production

    Scaffolded the infrastructure-as-code, the event handlers, encrypted secrets and network connectivity, structured logging and metrics, the operational-readiness review, and the rollout runbook. The whole lifecycle, not a feature on someone else's platform.

  2. 02

    Judgment

    Headed off a million-request-a-day problem before it shipped

    Caught at design time that his new service would hammer another team's API ~1M+ times a day for data it didn't need. He quantified the risk, proposed a leaner endpoint, drove the cross-team redesign to agreement, and held his own rollout until it was resolved.

  3. 03

    Reliability

    Drove production alerts from 258 to ~54 a month

    By fixing root causes and deleting dead code, not muting alarms: SQL retry/backoff, defunct infrastructure removed, thresholds tuned. Investigated every alert on-call and hit two consecutive zero-alert days.

  4. 04

    Influence

    AI-native engineering that spread beyond his team

    A division-wide talk on treating AI as a contractor that needs context: knowledge bases, automated workflows, and subagents that run multi-source incident investigations. His manager picked up one of his tools, a partner-team lead asked for his knowledge-base template to build their own, and he mentors an engineer on another team.

By the numbers

8months at Xero
258→54monthly production alerts, on his watch
48K→7Kbank records deduped in a 3-day build
1service shipped, empty repo to production

Top 1% for AI-assisted development, across Xero's 1,300+ engineers.

And within my 38-engineer domain

  • 0 reverts across 93 merged pull requests#1 of 38
  • Shipped to 29 repositories#1 of 38
  • Used AI tools 210 days, nearly every workday#1 of 38
  • 170+ story points deliveredTop quartile
Bank Feeds domain · 38 engineers · Oct 2025 to May 2026

What people say

In my team, you are the most prolific Claude user.
An SVP of engineering
Thank you Moravy for being my AI mentor. I've learnt so much, and I'll keep learning more.
A mentee, on another team
That's awesome!!! Thanks for sharing it.
His engineering manager, on a tool he then adopted
Your repo is so comprehensive I literally don't know how you get there. It's simply amazing.
An engineer he paired with

Writing, on building with AI

  1. 01how I use Claude Code day to day, six months laterSix months ago I wrote about how I use Claude Code day to day. One month, 101 sessions: investigations, feature work, code review, one ticket at a time. That post is still true. It's just no longer the whole picture.Read ↗
  2. 02How AI Agents Are Shifting the Software Development LifecycleRequirements. Design. Implementation. Code review. Testing. Deployment. Maintenance.Read ↗
  3. 03how I use Claude Code day to dayHere's a snapshot of what I've been doing with Claude Code over the last month across **101 sessions**:Read ↗

All writing →