This category is designed for anyone looking to run AI models on their own machines or servers without relying on the cloud. Through practical examples, I cover topics like local language models with Ollama, image generation workflows with ComfyUI, structured output and validation, and internal knowledge bases using RAG.
My notes don't just showcase successful results—I also share measured data on quantization comparisons, memory limits, and quality trade-offs. This helps you decide which model makes the most sense for your hardware and specific use case.
Articles in this section are written for both developers and non-technical teams. While diving into implementation details like model selection, prompt engineering, output validation with JSON schemas, and batch processing queues, I also address key questions for decision-makers around cost, privacy, maintenance, and team adoption. You'll also find anonymized case studies from my local image generation projects for textile and e-commerce brands.
If you are planning an AI solution where your data never leaves your company, take a look at my Desktop Applications & Local AI services.