ResellReel Video Learning
AI Engineer Full Course in 3 Hours [FREE]
Video details
About this video
2:53:55Length
108,928Views
2,552Likes
This is a full course on how modern AI systems actually work, and instead of jumping between disconnected examples, we learn everything by building one imaginary agent from start to finish. It's a refund agent for Cloud X Berry, a made-up online store, and it starts out simple: a customer says their order arrived broken and asks for a refund. By the end of the course that same agent is calling tools, pulling answers from company documents, running on real infrastructure, getting tested and red-teamed, and handling live traffic in production.
Every time the agent hits a wall, we stop and learn the idea that gets it past that wall. When it can only reply but can't actually do anything, we look at what makes something an agent and how tool calling works. When it needs to reach other systems, we cover MCP, agent skills, and AI gateways. When it doesn't know your company's policies, we get into RAG, knowledge graphs, and how to keep a knowledge base from going stale. From there we look at where a model's behavior comes from and how teams change it with fine-tuning, LoRA, distillation, and reinforcement learning, what kind of hardware all of this runs on and what it really costs, and finally what it takes to keep an AI system working once it's live.
You don't need a machine learning background for this. If you're a developer, an engineer, a product person, or just someone who wants to understand what's going on under the hood of the AI tools everyone's talking about, this course is meant for you. Everything is explained visually, one step at a time, and each idea builds on the one before it.
Chapters
0:00:00 Course Intro
0:02:38 Module 1 Intro: From Answering to Doing
0:03:26 1. What Makes Something an Agent
0:06:08 2. Five Core Subsystems Around an Agent: Perception, Reasoning, Action, Learning, Communication
0:09:40 3. Single Agent vs. Multi-Agent: When One Isn't Enough
0:13:09 4. Agent Memory: Why Agents Forget
0:16:53 5. The Context Window as a Budget: The "Lost in the Middle" Problem
0:21:05 Module 1 Recap: The Shape of an Agent
0:22:01 Module 2 Intro: From One Agent to a System
0:22:47 6. Tool Calling: How an Agent Does Something
0:28:20 7. MCP (Model Context Protocol): A Standard Way for Agents to Reach Tools
0:33:13 8. MCP vs. Agent Skills: Two Different Answers to the Same Problem
0:37:54 9. Agent Roles and Boundaries: Why Multi-Agent Systems Need Clear Ownership
0:43:01 10. The AI Gateway: A Different Front Door for AI Traffic Than a Website Needs
0:47:46 Module 2 Recap: Where Every Request Gets Checked
0:49:07 Module 3 Intro: Giving an Agent Something to Stand On
0:50:00 11. Beyond Basic RAG: Why Simple Retrieval Stops Working
0:54:53 12. A Field Guide to RAG Variants: When Simple, Self-Correcting, or Agentic RAG Applies
1:00:03 13. Knowledge Graphs vs. GraphRAG vs. Graph Engineering: Three Different Jobs
1:05:34 14. Agentic RAG: When Retrieval Becomes a Decision Instead of a Step
1:10:14 15. Building a Living Knowledge Base Instead of a Notes Graveyard
1:15:44 Module 3 Recap: When the Answer Lives in a Document
1:17:07 Module 4 Intro: Where a Model's Behavior Comes From
1:17:56 16. Data Architecture for AI: Pipelines, Versioning, and Keeping Training Data Honest
1:23:31 17. From Early Neural Networks to Transformers: Why Today's Models Look This Way
1:29:19 18. Changing a Model: Knowledge, Behavior, or Size?
1:34:18 19. LoRA: How Teams Fine-Tune Without Retraining Everything
1:38:59 20. Distillation: Teaching a Small Model to Act Like a Big One
1:43:37 21. Reinforcement Learning's Role in Modern Model Training
1:49:08 Module 4 Recap: Where Behavior Comes From, and How to Change It
1:50:29 Module 5 Intro: What It Takes to Run These Systems
1:51:19 22. The Chip Alphabet: CPU, GPU, TPU, NPU, and Why the Difference Matters
1:55:54 23. What CUDA Does, and Why AI Software Depends on It
2:00:42 24. Inside a GPU: What's Happening During Training vs. Inference
2:05:47 25. Clouds Built for AI: Why "Cloud" Isn't One Thing Anymore
2:10:59 26. Prompt Caching and Other Cost Levers Most Teams Miss
2:16:19 Module 5 Recap: What Runs Underneath, and What It Costs
2:17:42 Module 6 Intro: How You Know It Works, and Keeps Working
2:18:38 27. SDLC vs. AIDLC: Why the Development Lifecycle Itself Has to Change
2:23:16 28. DevOps vs. MLOps vs. LLMOps, and Tracing Every Request
2:27:42 29. AI Evals: How Production Teams Know a System Still Works
2:32:59 30. Why AI Projects Fail in Production
2:37:21 31. AI Red Teaming: Testing a System Before Someone Else Does
2:42:17 32. Platform Engineering for AI: Building the Internal Roads Other Teams Drive On
2:47:06 33. Putting It Together: The Anatomy of a Production-Ready AI System
2:52:35 Module 6 Recap: The Whole System, End to End
If this helped you make sense of how AI systems fit together, consider subscribing. I'd also love to hear in the comments which part clicked for you, or what you'd like me to go deeper on next.
#ArtificialIntelligence #AIAgents #MachineLearning #RAG
Every time the agent hits a wall, we stop and learn the idea that gets it past that wall. When it can only reply but can't actually do anything, we look at what makes something an agent and how tool calling works. When it needs to reach other systems, we cover MCP, agent skills, and AI gateways. When it doesn't know your company's policies, we get into RAG, knowledge graphs, and how to keep a knowledge base from going stale. From there we look at where a model's behavior comes from and how teams change it with fine-tuning, LoRA, distillation, and reinforcement learning, what kind of hardware all of this runs on and what it really costs, and finally what it takes to keep an AI system working once it's live.
You don't need a machine learning background for this. If you're a developer, an engineer, a product person, or just someone who wants to understand what's going on under the hood of the AI tools everyone's talking about, this course is meant for you. Everything is explained visually, one step at a time, and each idea builds on the one before it.
Chapters
0:00:00 Course Intro
0:02:38 Module 1 Intro: From Answering to Doing
0:03:26 1. What Makes Something an Agent
0:06:08 2. Five Core Subsystems Around an Agent: Perception, Reasoning, Action, Learning, Communication
0:09:40 3. Single Agent vs. Multi-Agent: When One Isn't Enough
0:13:09 4. Agent Memory: Why Agents Forget
0:16:53 5. The Context Window as a Budget: The "Lost in the Middle" Problem
0:21:05 Module 1 Recap: The Shape of an Agent
0:22:01 Module 2 Intro: From One Agent to a System
0:22:47 6. Tool Calling: How an Agent Does Something
0:28:20 7. MCP (Model Context Protocol): A Standard Way for Agents to Reach Tools
0:33:13 8. MCP vs. Agent Skills: Two Different Answers to the Same Problem
0:37:54 9. Agent Roles and Boundaries: Why Multi-Agent Systems Need Clear Ownership
0:43:01 10. The AI Gateway: A Different Front Door for AI Traffic Than a Website Needs
0:47:46 Module 2 Recap: Where Every Request Gets Checked
0:49:07 Module 3 Intro: Giving an Agent Something to Stand On
0:50:00 11. Beyond Basic RAG: Why Simple Retrieval Stops Working
0:54:53 12. A Field Guide to RAG Variants: When Simple, Self-Correcting, or Agentic RAG Applies
1:00:03 13. Knowledge Graphs vs. GraphRAG vs. Graph Engineering: Three Different Jobs
1:05:34 14. Agentic RAG: When Retrieval Becomes a Decision Instead of a Step
1:10:14 15. Building a Living Knowledge Base Instead of a Notes Graveyard
1:15:44 Module 3 Recap: When the Answer Lives in a Document
1:17:07 Module 4 Intro: Where a Model's Behavior Comes From
1:17:56 16. Data Architecture for AI: Pipelines, Versioning, and Keeping Training Data Honest
1:23:31 17. From Early Neural Networks to Transformers: Why Today's Models Look This Way
1:29:19 18. Changing a Model: Knowledge, Behavior, or Size?
1:34:18 19. LoRA: How Teams Fine-Tune Without Retraining Everything
1:38:59 20. Distillation: Teaching a Small Model to Act Like a Big One
1:43:37 21. Reinforcement Learning's Role in Modern Model Training
1:49:08 Module 4 Recap: Where Behavior Comes From, and How to Change It
1:50:29 Module 5 Intro: What It Takes to Run These Systems
1:51:19 22. The Chip Alphabet: CPU, GPU, TPU, NPU, and Why the Difference Matters
1:55:54 23. What CUDA Does, and Why AI Software Depends on It
2:00:42 24. Inside a GPU: What's Happening During Training vs. Inference
2:05:47 25. Clouds Built for AI: Why "Cloud" Isn't One Thing Anymore
2:10:59 26. Prompt Caching and Other Cost Levers Most Teams Miss
2:16:19 Module 5 Recap: What Runs Underneath, and What It Costs
2:17:42 Module 6 Intro: How You Know It Works, and Keeps Working
2:18:38 27. SDLC vs. AIDLC: Why the Development Lifecycle Itself Has to Change
2:23:16 28. DevOps vs. MLOps vs. LLMOps, and Tracing Every Request
2:27:42 29. AI Evals: How Production Teams Know a System Still Works
2:32:59 30. Why AI Projects Fail in Production
2:37:21 31. AI Red Teaming: Testing a System Before Someone Else Does
2:42:17 32. Platform Engineering for AI: Building the Internal Roads Other Teams Drive On
2:47:06 33. Putting It Together: The Anatomy of a Production-Ready AI System
2:52:35 Module 6 Recap: The Whole System, End to End
If this helped you make sense of how AI systems fit together, consider subscribing. I'd also love to hear in the comments which part clicked for you, or what you'd like me to go deeper on next.
#ArtificialIntelligence #AIAgents #MachineLearning #RAG
Explore related learning
Video topics
AI engineering AI agents how AI systems work AI full course agentic AI tool calling MCP Model Context Protocol agent skills multi-agent systems AI gateway RAG retrieval augmented generation agentic RAG GraphRAG knowledge graphs vector database LLM LoRA distillation reinforcement learning GPU TPU NPU CUDA AI infrastructure prompt caching AI evals LLMOps MLOps AI red teaming platform engineering AI in production context window AI course for beginners