Video course · 9 chapters · 22 min · certificate
Build AI Agents: A Practical Guide to Architecture & Tools
Understand what AI agents really are and pick the right way to build one: the building blocks, then four tiers from no-code platforms to low-code, agent harnesses and full code.
What you'll learn
- Explain what an AI agent actually is
- Name the building blocks: model, tools, loop, memory
- Compare no-code, low-code, harness and full-code options
- Understand the agentic loop in code
- Choose the right tier for a use case
- Plan how to learn agent building
Chapters
9 chapters · 21:34-
1:53
01Basics Members
What agents really are
Cutting through the hype: agents aren't magic brains, they're models plus tools in a loop.
-
5:22
02Architecture Members
The building blocks
The LLM is the brain that predicts text; tools, a loop and memory make it act.
-
1:10
03Landscape Members
The four tiers
Four ways to build agents, based on how much of the agent you build yourself.
-
1:27
04Learning Members
How to get good at this
Watching videos isn't enough: learn agents by structured practice and building.
-
2:44
05No-code Members
Tier 1: no-code
Platforms like Lindy, Gumloop, Stack AI and Relevance AI handle everything; you configure.
-
2:26
06Low-code Members
Tier 2: low-code
Drag-and-drop editors with code blocks, for real logic without managing a codebase.
-
2:42
07Harness Members
Tier 3: agent harnesses
Harnesses like OpenClaw, Hermes or Letta provide the agentic loop; you add tools and config.
-
2:10
08Full code Members
Tier 4: full code
Building the loop yourself: system prompt, tools, and the model calling tools until done.
-
1:40
09Decision Members
How to pick a tier
Choosing no-code, low-code, harness or full code based on logic, integrations and deployment needs.
Jobs this course suits
Our AI checked this course against 500 jobs; these get the most out of it. Each job links to its learning path.