| | | 📝 A Few Words | | | | | | I wrote a book so you can stop renting your AI
Last August, OpenAI retired GPT-4o overnight and moved everyone to GPT-5. People who had built their daily work around a fast, predictable model woke up to a slower one that behaved differently, with no way back. This June, a US export control directive forced Anthropic to cut Fable 5 and Mythos 5 for every customer at once. Teams lost the models they were building on in an afternoon, for reasons that had nothing to do with their own work.
None of those people did anything wrong. They just did not own what they ran. The model, the price, and the rules sat in someone else's hands, and any of the three could move without warning.
My new book exists because I got tired of that arrangement.
Local AI Engineering with Ollama: Run, understand, customize, fine-tune, and build agentic apps on your own hardware.
This is a practical book, not a survey of the field. With no history lessons or predictions about where AI is headed. You install the runtime, pull a model, and by the end you have three things sitting on your own disk:
- A custom model packaged with a Modelfile that does one job the same way every time, that a teammate can pull and run with zero setup.
- A fine-tuned model trained with QLoRA and Unsloth, then exported to GGUF and run in Ollama.
- A chat application you build in nine passes until it becomes an advanced agent, with conversation history, streaming, context trimming, LangChain summarization, Redis caching, mem0 long-term memory, function calling, and tools served over MCP.
- And more!
Along the way you learn what a model is actually doing (tokens, weights, embeddings, the KV cache, quantization), how to size a model against your RAM or VRAM before you download it, how to drive Ollama from its HTTP API, and how to control the context window so the model stops silently forgetting where a long chat started.
The stack you practice on: Ollama, Unsloth, LangChain, Redis, Docker, mem0, and Open WebUI.
Who it's for
If you can run a command and edit a file, you are qualified. No ML degree required, and none wanted. This is the book I needed when I started, written for the developer in the middle, past the marketing pages and short of the research papers.
What makes it different
Every command in the book was run on a real machine. Every output you see, the JSON responses, the error messages, the token counts, the training logs, came from an actual session, not from docs I trusted and pasted in. When Ollama behaved differently from its own documentation, I say so and pin the version it happened on...etc Where accuracy and polish pulled apart, accuracy won. That is the part that ages well.
28 modules, 91 sections, lifetime access and updates, a built-in AI assistant (SenseiOne) for your questions, and a 30-day money-back guarantee.
Get your copy
👉 On FAUN.sensei: Local AI Engineering with Ollama. Use code OLLAMA20 at checkout for 20% off. The code expires July 8, 2026 at 11:59 PM, so move before then.
👉 On Amazon.com: paperback and Kindle editions are live here (also available in the other marketplaces: .fr .de ..etc)
👑 If you want to stop renting and start owning, this is for you. Get a model running tonight, and keep going.
The rest follows from there. |
| | | | |
|
|
