Technical articles exploring real engineering problems and how I solved them. These deep dives cover architecture decisions, performance optimizations, deployment strategies, and the technical details that actually matter when building software people pay for.
No basic tutorials. No rehashed documentation. Just honest reviews of what worked, what failed, and what I learned building B2B/B2C tools as an independent developer.
Topics include web development, AI integration, video processing, automation tools, database optimization, API design, and the boring infrastructure work that makes everything else possible. Filter by tag below to find what you are looking for.
Showing 10 articles
A production benchmark comparing Claude Code (CLI) vs. Google Antigravity (Gemini 3 Pro). Why I switched workflows after hitting the Opus 4.5 limits, despite the "vibe coding" flaws.
Why deploying open-source TTS models like Kokoro-82M on embedded systems faces GPLv3 licensing traps and Python runtime costs.
Switching from html-to-image to Canvas rendering achieved 5x faster video exports. Technical breakdown and code architecture.
Real data from marketing a SaaS product on Reddit. CTR metrics, analytics setup, ICP definition, and whether Reddit marketing works.
MindLock.io gives you data sovereignty with local AI distillation and unlimited storage. Break free from AI vendor lock-in.
Learn what ChatGPT, Claude, and AI tools store about you, why they do it, and how to manage your AI memory effectively.
Build a persistence system using distilled conversations that makes your AI brutally honest and accountable.
Use hooks and custom agents to enforce best practices, prevent AI slop, and build production-ready applications.
Train AI models to stop sugarcoating, build persistent behavior, and boost productivity using agents, Claude Code, and MCP servers.
Build a complete self-hosted infrastructure with Raspberry Pi 5, Ubuntu Server, and Docker. Run Gitea, n8n, Docmost on your own hardware.