AI Agent or n8n Workflow? Choose Determinism Before Autonomy
Use deterministic workflow automation for predictable work and AI agents for ambiguous decisions. A practical framework for choosing where each belongs.
Useful software starts with clear thinking.
Practical engineering studies, intelligent systems, and open software projects by Ray. Built with clarity, zero bloat, and verified evidence.
Technical writing
Recent Technical Articles
In-depth engineering notes, multi-agent review workflows, and developer infrastructure explained through working examples.
Always-On AI Agent Architecture
Explore the structural patterns, tool permissions, and human approval gates required to build reliable, always-on AI agent architectures using GitHub Actions and automated backends.
Debugging R8 Release-Only Crashes Without Disabling Optimization
A systematic way to diagnose Android crashes that appear only after R8 optimization, from retracing stack traces to writing the narrowest keep rule that fixes the real boundary.
Featured software
Independent Tools
Focused software and productivity tools designed and engineered by RayLabs.
Zero-subscription Obsidian vault synchronization using personal GitHub repositories with automatic 3-way merge conflict safety.
Monumental presentation clock designed for speakers and conferences, featuring zero-jitter tabular numerals, ambient audio, and full-screen projection.
Open-source projects
Recent GitHub Projects
Public repositories and active software tools created by Ray across Android, web, and automation.
Android Base Prime
AndroidA modular, high-performance base utility architecture for modern Android and Kotlin applications.
Harmonia AI Health
AI PLATFORMAI-powered platform for comprehensive health analysis, skin cancer detection, and personal diagnostics.
WeeabooHub
APPCurated high-resolution anime digital artwork and wallpaper curation application.
External publications
Also Published on Medium
RayLabs publishes canonical technical notes here first, then syndicates select articles to the wider Medium community.
SYNDICATED
Multi-Agent Review Pipeline for AI Coding Agents
A multi-agent review pipeline for AI coding work separates implementation, independent review, final audit, and human approval.
Read on Medium
Feature story
Should Your AI Coding Agent Run on a VPS or Your Mac?
Choose where an AI coding agent should run by separating interactive development, always-on orchestration, and heavy build workloads.
Read on Medium
Medium
The Ultimate Guide to Product Management: PRDs, Roadmaps, GTM, and Metrics
A Product Manager can spend one hour discussing a user story, another hour reviewing a roadmap, and looking at metrics.
Read on Medium