Slate
Explore
Tracks
History
Pricing
Start session
All tracks
Computer Science
AI agents, pragmatically
Tool use, planning, and loops — how LLM agents actually work and why most of them fail.
6 lessons
~100 min total
First principles
What you'll learn
Build a working mental model of what an 'agent' actually is
Understand tool use, planning, and the role of memory
Recognize the failure modes that kill most agent projects
Progress
0 / 6
Track complete ✓
Lessons
1
What an 'agent' even means
Cutting through the hype to a useful definition.
3 objectives
2
Tool use, the core capability
Structured output, function calling, and why it's harder than it looks.
3 objectives
3
Planning vs reacting
ReAct, plan-and-execute, and the space between.
3 objectives
4
Memory: short, long, and selective
The three memory layers most teams reinvent badly.
3 objectives
5
How you know it's working
Evals for agents — harder than for models, just as important.
3 objectives
6
Why most agents fail
Loops, tool misuse, compounding errors — and what actually helps.
3 objectives
Related tracks
How LLMs actually work
From tokens to transformers to the training loop — what's really happening when a model responds.
7 lessons
SLATE
Software architecture
Why real systems are shaped the way they are — monoliths, services, databases, caches, queues, and the tradeoffs behind each.
8 lessons
SLATE
How the internet actually works
What happens between typing a URL and the page appearing — packets, DNS, TCP, TLS, the works.
8 lessons
SLATE