I'm a senior backend engineer working remotely from Canada. Day to day I deal with: millions of device messages, real-time alerting, distributed system reliability — and lately, what excites me most: redesigning internal operations with AI agents.
Over the past few years I've built IoT device alerting systems, led an MQTT v5 migration, redesigned an ELK indexing strategy, and most recently wired an agentic AI chatbot into an internal RAG knowledge base — so IT admins can query, troubleshoot, and execute management operations in natural language.
This site is where I turn those experiences into judgment. I write for two kinds of readers: engineers like me, and anyone trying to build with AI without being dragged around by hype. AI can generate many options; the hard part is deciding what is worth keeping and what is safe to ship.
What I'm up to now
- Building AI infra — wrangling Kafka, Go, and vector databases daily
- Writing the "Classics Are Dead" series — one essay a month
- Putting together System Design learning notes
Current projects
✍️ zoe-site this site
A personal site built with Astro + Notion to publish notes on AI practice, backend systems, and product judgment.
🧪 practice-memory Local-first AI practice memory app
A local-first, cost-aware desktop app for self-learners to manage practice and spaced review.
🛠️ AI Infra at work
Designing agentic AI chatbot backends, RAG over internal documentation, LLM-as-judge evaluation, and permission-gated tool execution.
What I've worked on
2021 — Now Senior Back-end Developer
Designed IoT device alerting (100+ enterprise customers / 100K+ devices), real-time tracking, led MQTT v5 migration and ELK redesign. Currently building an agentic AI chatbot backend.
2018 — 2019 Systems Automation Engineer · Applied Optoelectronics
2016 — 2017 Manufacturing Automation Engineer · TSMC
Education
🐊 University of Florida
Certifications
✓ AWS Certified Generative AI Practitioner