Fundamentals
8 modules · ~3 h 15 min
--/8 completed
Showing 8 of 33
Overview
Cards
Fundamentals
Fundamentals01
LLM Fundamentals
Description
What a language model actually does, what the context window costs you, and which knobs change its output.
- Length
- 2,498 W · 25 MIN
- Sources
- 11
- Requires
- Nothing
Fundamentals02
Training LLMs
Description
How a general model is born in pre-training, and the two ways you specialise it afterwards.
- Length
- 2,637 W · 25 MIN
- Sources
- 7
- Requires
- 1
Fundamentals03
RAG & Embeddings
Description
Giving a model knowledge it was never trained on, using embeddings and a vector store.
- Length
- 2,281 W · 30 MIN
- Sources
- 14
- Requires
- 1, 2
Fundamentals04
Tool Calling
Description
How a model reaches outside itself, the mechanism underneath every agent.
- Length
- 1,790 W · 25 MIN
- Sources
- 1
- Requires
- 1
Fundamentals05
Memory
Description
Three kinds of memory, parametric, working and long-term, and where each one actually lives.
- Length
- 1,356 W · 20 MIN
- Sources
- 1
- Requires
- 1, 3
Fundamentals06
AI Agents
Description
The agent loop, and what turns a single model call into multi-step work.
- Length
- 1,639 W · 25 MIN
- Sources
- 7
- Requires
- 4, 5
Fundamentals07
Multi-Agent Systems
Description
When one agent stops being enough, the two architectures to know, and the two problems you inherit.
- Length
- 1,394 W · 25 MIN
- Sources
- 3
- Requires
- 6
Fundamentals08
Observability
Description
Traces are the only record of what an agent actually did, and the loop that turns them into a better agent.
- Length
- 1,469 W · 20 MIN
- Sources
- 6
- Requires
- 6
Table
| # | Module | Level | Description | Length | Sources | Completion | Requirements |
|---|---|---|---|---|---|---|---|
| 01 | LLM FundamentalsREADY | Fundamentals | DescriptionWhat a language model actually does, what the context window costs you, and which knobs change its output. | 2,498 W · 25 MIN | 11 | — | |
| 02 | Training LLMsREADY | Fundamentals | DescriptionHow a general model is born in pre-training, and the two ways you specialise it afterwards. | 2,637 W · 25 MIN | 7 | 1 | |
| 03 | RAG & EmbeddingsREADY | Fundamentals | DescriptionGiving a model knowledge it was never trained on, using embeddings and a vector store. | 2,281 W · 30 MIN | 14 | 1, 2 | |
| 04 | Tool CallingREADY | Fundamentals | DescriptionHow a model reaches outside itself, the mechanism underneath every agent. | 1,790 W · 25 MIN | 1 | 1 | |
| 05 | MemoryREADY | Fundamentals | DescriptionThree kinds of memory, parametric, working and long-term, and where each one actually lives. | 1,356 W · 20 MIN | 1 | 1, 3 | |
| 06 | AI AgentsREADY | Fundamentals | DescriptionThe agent loop, and what turns a single model call into multi-step work. | 1,639 W · 25 MIN | 7 | 4, 5 | |
| 07 | Multi-Agent SystemsREADY | Fundamentals | DescriptionWhen one agent stops being enough, the two architectures to know, and the two problems you inherit. | 1,394 W · 25 MIN | 3 | 6 | |
| 08 | ObservabilityREADY | Fundamentals | DescriptionTraces are the only record of what an agent actually did, and the loop that turns them into a better agent. | 1,469 W · 20 MIN | 6 | 6 |
General notes
The starting point of the series: the core concepts every AI builder needs before going further.