Jun 3
2026
engineering
llmself-hostedmac-studio

How Claude Code and I picked my local coding model (and the rubric that lied to me)

I asked Claude Code to help me figure out which local LLM was actually best for coding on my Mac Studio. It refused to guess and built a benchmark instead. Here is what we found, what surprised me, and where I had to override Claude.

>
Jun 2
2026
ai-tools
local-modelsllmollama

Local AI Models: A Plain-English Guide to the Jargon

bf16, Q4_K_M, GGUF, MTP — model names are a wall of acronyms. Here's what they actually mean and whether the model will run on your Mac.

>
May 29
2026
engineering
dockercloudflare-tunnelsself-hosting

How to Expose Your Docker Apps to the Internet with Cloudflare Tunnels

The complete guide to running production Docker apps on a local machine with Cloudflare Tunnels — no open ports, no static IP, no VPS needed.

>
May 25
2026
engineering
llmself-hostedlinux

How to Run llama.cpp on a Mac Pro 6,1 with Dual AMD FirePro D300 GPUs on Ubuntu

Running 3B LLMs at 22 tok/s on a 2013 Mac Pro with dual FirePro D300 GPUs under Ubuntu. Covers Vulkan setup, llama.cpp build, systemd config, and three hardware-specific traps that silently break inference.

>
May 20
2026
automation
homebridgehomekitsmart-home

How I Connected a Cooper & Hunter Mini-Split to Apple Home (Without the Official App)

How to identify Cooper & Hunter mini-splits as Midea OEM, extract credentials from the macOS companion app, and connect them to Apple Home via homebridge-midea-platform.

>
May 6
2026
marketing
marketing-opsartificial-intelligencen8n

Marketing Ops: The Only Job AI Can't \"Execute\" Away

AI writes copy and wires automations in minutes. But the person who connects pricing to CRM to dashboard is about to become the most secure job in marketing. Here's why, with a real system I built.

>