About MasterAIKit

I'm building and testing local AI systems on real hardware so you don't have to waste weeks guessing.

Why MasterAIKit Exists

Most local AI content online is garbage. It's either:

• Generic tutorials copied from docs with no real testing
• Hype videos showing "amazing" results without benchmarks or error logs
• Outdated guides that don't reflect current tool versions
• Too technical — written for ML engineers, not normal PC users

MasterAIKit is the opposite. Every guide is:

• Tested on a real machine — with specs, versions, and dates listed
• Documented honestly — including what broke and how I fixed it
• Screenshot-backed — not stock images, actual captures
• Written for real people — clear steps, no jargon dumps

The Test Bench

This rig gives me 24GB total VRAM across two GPUs — enough to test everything from small 3B models up to 32B+ models with quantization. Every benchmark, screenshot, and error log on this site comes from this machine.

What's Different About These Guides

What I Use Local AI For

• Research & writing — using Goose as a research agent with local models
• Code assistance — local coding agents for privacy-sensitive projects
• Document chat — RAG pipelines to chat with PDFs and notes privately
• Obsidian integration — using my note vault as AI agent memory
• Content creation — this entire site is built with Goose as a research/planning partner

Contact me →