Silver Tree AI Trilogy · Part 2
The Local LLM Bible — Beyond the Limits of Subscription AI
For when the answer is not ChatGPT but a model your company owns.
- PUBLISHERSilver Tree Books
- PUBLISHED5 January 2026
- ISBN979-11-993543-7-1
- EXTENT602 pages · 194 × 260 mm
- CATEGORYArtificial intelligence · Engineering
About the Book
This book begins where most AI adoption stalls — the moment someone pastes company data into a subscription chatbot and hesitates. It answers three needs that hosted models cannot: data sovereignty, a model that knows your internal rules and vocabulary, and independence from a single vendor.
The whole lifecycle is here in one volume: local infrastructure on Ollama, prompt and role design, RAG to connect your own documents, and fine-tuning to teach the model your house style. Two build-outs run end to end — an internal knowledge assistant and a customer-support assistant — from proof of concept to production, with runnable commands, Modelfiles, and architecture diagrams. It favors systems you can stand up with the resources you already have over theory you cannot deploy.
Contents
- PrologueWhy local: general-purpose LLMs vs a model that is yours
- 1LLMs and Ollama — tokens, embeddings, context; transformers; how prompting, RAG, and fine-tuning relate
- 2Building local infrastructure on Ollama — design and install, model selection, Modelfile basics
- 3Prompt and role design — principles, system prompts, task patterns, advanced technique
- 4RAG: connecting company data — architecture, ingestion, chunking and embedding, vector search, response quality
- 5Fine-tuning — dataset design, quality control, LoRA/QLoRA, deploying to Ollama
- 6Two builds — internal knowledge base and customer support, PoC to production
- Afterword
About the Author
Kim writes the bridge between AI in practice and AI in principle — a trilogy running from prompt engineering to enterprise local LLMs to the mechanics underneath.