# Learn AI Visually: How LLMs Work, Prompt Engineering & AI Agents | Koso Learn > A free, visual guide to modern AI. Three chapters: how the language engine inside ChatGPT and Claude works, prompt engineering techniques that hold up in production, and how AI agents are built. A diagram in every single section. [](/) [Services](/#services)[AI OS](/ai-os)[Training](/training)[Education](/education)[Agents](/agents)[Learn](/learn) … [Book a call ](/book) 3 chapters, free to read # Learn AI. Visually. Most guides to this subject are long stretches of text about something that's easier to picture than to describe. This one leans the other way: three chapters, each one built around diagrams rather than paragraphs of dense explanation. No math background needed. [Start chapter 01 ](/learn/how-llms-work)[Browse the chapters](#chapters) The whole guide, in one picture youno math needed01 · the enginehow LLMs worktokens & embeddingsattention & trainingcontext & hallucination02 · promptingsteering the enginefew-shot & CoTstructured outputgrounding & debugging03 · agentsengines that actthe loop & toolsRAG & multi-agentguardrails & evalsshipping Swipe the diagram sideways to see it all. The path through the guide. Each chapter leans on ideas from the one before it: the engine explains why prompting works, and prompting explains how agents are steered. Even so, every chapter can be read on its own. The chapters ## Engine. Prompts. Agents. Read them in order, since each one leans on ideas from the last. Or jump straight to whichever chapter you need. Every section pairs plain text with a diagram, drawn in the same style throughout. 01Foundations ### How LLMs Work 9 sections \~25 min [Read the chapter ](/learn/how-llms-work) What's actually inside ChatGPT, Claude and Gemini. Tokens, embeddings, attention, training, temperature, the context window, and why models sometimes make things up. Every idea drawn out. 1. [01One job: predict the next token](/learn/how-llms-work#prediction) 2. [02Tokens: how text becomes numbers](/learn/how-llms-work#tokens) 3. [03Embeddings: meaning becomes a map](/learn/how-llms-work#embeddings) 4. [04Attention: every token reads every other](/learn/how-llms-work#attention) 5. [05The transformer stack](/learn/how-llms-work#stack) 6. [06Training: from internet to assistant](/learn/how-llms-work#training) 7. [07Temperature: the dial on the dice](/learn/how-llms-work#sampling) 8. [08The context window: working memory](/learn/how-llms-work#context) 9. [09Hallucination: confident, not correct](/learn/how-llms-work#hallucination) 02Practitioner ### Prompt Engineering 8 sections \~22 min [Read the chapter ](/learn/prompt-engineering) The techniques that reliably change model output: prompt anatomy, few-shot examples, chain of thought, self-consistency, structured output, the system-prompt stack, grounding and a debugging loop. 1. [01The anatomy of a strong prompt](/learn/prompt-engineering#anatomy) 2. [02Zero-shot, one-shot, few-shot](/learn/prompt-engineering#shots) 3. [03Chain of thought: make it think out loud](/learn/prompt-engineering#chain-of-thought) 4. [04Self-consistency: ask many, take the vote](/learn/prompt-engineering#self-consistency) 5. [05Structured output: prompts that feed software](/learn/prompt-engineering#structured-output) 6. [06The prompt stack: system, developer, user](/learn/prompt-engineering#prompt-stack) 7. [07Grounding: bring your own facts](/learn/prompt-engineering#grounding) 8. [08Debugging prompts like an engineer](/learn/prompt-engineering#debugging) 03Builder ### AI Agents 8 sections \~26 min [Read the chapter ](/learn/ai-agents) From chat to systems that do work. The agent loop, tool calling, memory, planning patterns, multi-agent teams, RAG pipelines, guardrails and evals, plus when an agent is the wrong answer. 1. [01The agent loop: think, act, observe](/learn/ai-agents#loop) 2. [02Tool use: hands for the model](/learn/ai-agents#tools) 3. [03Memory: what agents remember](/learn/ai-agents#memory) 4. [04Planning patterns: ReAct and plan-and-execute](/learn/ai-agents#planning) 5. [05Multi-agent teams](/learn/ai-agents#multi-agent) 6. [06RAG: retrieval-augmented generation](/learn/ai-agents#rag) 7. [07Guardrails, evals and humans in the loop](/learn/ai-agents#guardrails) 8. [08Agent or workflow? Choosing honestly](/learn/ai-agents#agent-or-workflow) How this guide works ### Every section is drawn Attention, temperature, the agent loop, RAG. Each concept gets its own diagram, because architecture is usually easier to follow as a picture than as a paragraph. ### Plain words, honest trade-offs No math, no hand-waving. Each technique comes with when it works, what it costs, and the failure mode you're likely to hit once you actually use it. ### Grounded in working examples The patterns in chapter 3 aren't just theory. A handful of [working agents](/agents) built this way are open to read through, if you want to see one in practice. A custom AI development agency. Bespoke agents, copilots and products engineered for teams that want a moat, not a subscription. Ideanab Private Ltd is a company registered in England and Wales, no. [16344048](https://find-and-update.company-information.service.gov.uk/company/16344048). Registered office: 124 City Road, London, EC1V 2NX, United Kingdom. ### Company * [Services](/#services) * [Custom AI Agency](/custom-ai-agency) * [AI Product Development](/ai-product-development) * [AI OS](/ai-os) * [AI Engineering](/ai-engineering-company) * [AI Consulting](/ai-consulting-company) * [AI Consulting Services](/artificial-intelligence-consulting-services) * [Training](/training) * [Education](/education) * [Free AI Agents](/agents) * [Learn AI](/learn) * [Fractional CAIO](/caio) * [About Us](/about) * [Careers](/careers) ### Contact * [Contact us](/contact) * [Book a call](/book) * [LinkedIn](https://www.linkedin.com/company/koso) * [Instagram](https://www.instagram.com/okaashish) ### Legal * [Terms of Service](/terms) * [Privacy Policy](/privacy) * [Cookie Policy](/cookies) * [Acceptable Use Policy](/acceptable-use) © 2026 Koso, a brand of Ideanab Private Ltd. All rights reserved. [Terms of Service](/terms)[Privacy Policy](/privacy)[Cookie Policy](/cookies)[Acceptable Use Policy](/acceptable-use) Crafted with care · Built to compound