Build a working mental model of LLMs, then map it onto AWS.

Module 01 of 16 · track Think · Estimated time: 3–4 hours · Prerequisites: Module 00.

Zero math. This module builds the intuition you need to make model choices you can defend — context windows, tokenisation, mixture-of-experts, temperature — and then connects each concept to the AWS surface where you will actually use it.

Open the module on GitHub ↗: the slides, notebooks, exercises and solutions are there. This page is the reading order, with the tutorial and the workbench alongside.

The lesson that frames this module is What Is the Agentic PDLC? The P0 to P3 Framework Explained (10 min): why software with a model in it needs a different lifecycle.

• Explain next-token prediction, context windows and tokenisation without hand-waving
• Read a model card and predict how the model will behave on your workload
• Choose between model families on latency, cost and capability — not vibes
• Map every LLM concept onto its Amazon Bedrock equivalent

Portable GenAI concepts	AWS-specific surface
Next-token prediction	Amazon Bedrock model catalogue
Tokenisation	Model IDs and regional prefixes
Context window	Cross-region inference profiles
Mixture-of-Experts
Temperature and sampling
Model families

Portable concepts transfer to any stack. The AWS column is where this module touches the cloud — see docs/concepts/portability-matrix.md.

#		Step	What it is
1	📖	slides/Day1.5_LLM_AWS_Bridge.pptx	The bridge deck — LLM concepts mapped to AWS
2	✏️	exercises/LLM_Intuition_Bank.md	Self-test bank: MCQs, diagrams, error-fixing. Closed-book first.
3	📊	activities/Day15_Companion_Workbook.xlsx	Companion workbook
4	✏️	exercises/Day1.5_Ad-Hoc_Exercise_PickTheModel.pdf	Pick the model — defend the choice

📖 read   💻 run   ✏️ practise   📊 workbook   🔖 reference

There is no separate solutions/ folder here because the answers travel with the questions: the LLM Intuition Bank carries its own answer key at the end. Attempt each section closed-book first — the italic Breadcrumb lines are nudges, not answers.

• Assuming a bigger context window is always better. It costs tokens and dilutes attention.
• Forgetting the us. / eu. inference-profile prefix on Bedrock model IDs — the single most common first error.

Every error anyone has hit in this course is in the Error Index, by the exact string. Model access and region problems are in Model and Region Notes. If it is a question rather than an error, check the answered ones first, then ask in Q&A.

Frameworks

• Token Tax Ledger — Where every token goes, and who pays

Quick reference

• Model selection — Four dimensions, the routing move, cost per resolved task
• Bedrock Converse API — The API you are about to meet

---

← Module 00 · Agentic Foundations · All modules · Module 02 · Amazon Bedrock Essentials →

_{✏️ This page is generated from modules/01-llm-and-aws-bridge/README.md. To change it, edit the module README: an edit made here is replaced at the next sync.}