AI Engineering Full Course | 8+ Hours | LLMs, RAG, LangChain, LangGraph, AI Agents
Description: Ladies and gentlemen, welcome aboard NeonLearn flight AI-10. 🚀
Our journey to becoming an AI Engineer includes 10 projects, 10 planets to explore, from building a chatbot from scratch to multi-agent systems.
🪐 1. CHATBOT: Build a streaming chatbot from scratch with the Groq API. Learn tokens and tokenizers, context management, conversation summarization, token budgets and cost estimation.
🪐 2. CLASSIFIER: Sentiment analysis with an LLM, plus a bonus: zero-shot classification with NLI models running locally via Hugging Face Transformers.
🪐 3. EMBEDDINGS: Turn text into vectors with Sentence Transformers, then measure meaning with cosine similarity and build semantic search.
🪐 4. RAG: Chat with your own PDFs. Learn chunking, embeddings, the Chroma vector store, grounded prompting and answers with source citations.
🪐 5. AGENT: Build AI agents with LangChain that use tools such as web search with Tavily and custom tools, and get real work done for you.
🪐 6. STRUCTURED: Get reliable structured output from LLMs, extracting clean data into Pydantic models with JSON schema.
🪐 7. MEMORY: Give your agents long-term memory with checkpointers and PostgreSQL, and keep separate conversation threads.
🪐 8. MIDDLEWARE: Control agent behavior with middleware, such as automatic conversation summarization.
🪐 9. LANGGRAPH: Design stateful workflows as graphs, such as a write → review → revise loop that improves until it's good enough.
🪐 10. MULTI-AGENT: Build a team of AI agents with the supervisor pattern, where one agent delegates to specialist agents.
📘 FLIGHT MANUAL (GitHub Repository)
All the course code is waiting for you on board:
https://github.com/mohamad-sw/ai-engineering-tutorial
⭐ If this journey helps you, please give the repo a star. It helps more passengers find this flight.
🧭 FLIGHT PLAN
🛫 BOARDING
00:00:00 – Course Introduction
00:04:27 – Setting Up the Development Environment
🪐 PLANET CHATBOT
00:13:35 – Building the Chat Function
00:18:10 – What is an LLM
00:26:51 – What is Hugging Face?
00:36:14 – Using the Groq Provider
00:41:43 – Using a Wrapper to Interact with LLMs
00:49:42 – Sending Requests to the LLM
01:04:31 – LLM Parameters
01:13:45 – Adding a Conversation Loop
01:18:26 – Conversation History
01:23:02 – Implementing Conversation History
01:26:20 – System Role
01:32:52 – What Are Tokens?
01:41:26 – Counting Total Tokens
01:48:01 – Making the Program Object-Oriented
01:57:45 – Calculating Cost
02:03:56 – Tiktoken
02:10:45 – Using Any Model's Tokenizer
02:19:49 – Streaming
02:30:31 – Context Window Management: The Sliding Window Strategy
02:44:06 – Summarizing Conversation History
🪐 PLANET CLASSIFIER
02:55:53 – Sentiment Analysis
03:03:39 – Prompt Engineering
03:10:23 – Natural Language Inference (NLI)
03:18:52 – BERT Model
03:25:10 – Using Pipeline to Run a Model
🪐 PLANET EMBEDDINGS
03:32:48 – What Are Embeddings?
03:40:43 – All-MiniLM-L6-v2
03:42:40 – Using Sentence Transformers Models
03:47:02 – Semantic Search
🪐 PLANET RAG
03:58:35 – Retrieval-Augmented Generation (RAG)
04:08:10 – The Retrieval Component
04:16:27 – Preparing the Knowledge Base
04:27:46 – Data Chunking
04:38:30 – Converting Chunks into Embeddings
04:46:09 – Using the Chroma Vector Database
04:59:29 – Generating IDs for Chunks
05:04:21 – Retrieving Chunks from the Vector Database
05:13:34 – Generation
🪐 PLANET AGENT
05:26:02 – What Is an AI Agent?
05:35:59 – Using the LangChain Wrapper
05:42:04 – Adding a Conversation Loop
05:48:24 – Building an Agent with create_agent Function
06:00:28 – Three Approaches to Building an Agent
06:04:37 – Building a Web Search Agent
06:11:20 – The MCP Protocol
🪐 PLANET STRUCTURED
06:18:43 – Dynamic Prompts
06:21:52 – The PromptTemplate Class
06:26:28 – The ChatPromptTemplate Class
06:35:15 – The MessagesPlaceholder Class
06:39:55 – Generating Structured Output
06:50:36 – Extracting Information with Structured Output
🪐 PLANET MEMORY
06:55:55 – Checkpointers
07:00:27 – Using InMemorySaver
07:06:28 – Switching Between Conversations
07:10:24 – Using PostgresSaver
07:18:18 – Reading Messages from the Database
🪐 PLANET MIDDLEWARE
07:22:32 – What Are Middlewares?
07:30:25 – Summarization Middleware
🪐 PLANET LANGGRAPH
07:35:03 – What Is LangGraph?
07:48:15 – Conditional Edges
07:55:42 – Automated Article Writing and Revision System
🪐 PLANET MULTI-AGENT
08:16:00 – Multi-Agent Systems
08:21:31 – The Supervisor Approach
🎟️ Boarding is free. If you know someone who dreams of becoming an AI Engineer, share this flight with them. Your captain will be grateful. 🙏
💬 Leave a message for the captain in the comments. He reads them all.
ℹ️ This course is not affiliated with or sponsored by any of the services or tools used in it.
Please fasten your seatbelts and keep your engines on. 🚀
#AIEngineering #LLM #RAG #AIAgents #LangChain #LangGraph #Python Close description
Our journey to becoming an AI Engineer includes 10 projects, 10 planets to explore, from building a chatbot from scratch to multi-agent systems.
🪐 1. CHATBOT: Build a streaming chatbot from scratch with the Groq API. Learn tokens and tokenizers, context management, conversation summarization, token budgets and cost estimation.
🪐 2. CLASSIFIER: Sentiment analysis with an LLM, plus a bonus: zero-shot classification with NLI models running locally via Hugging Face Transformers.
🪐 3. EMBEDDINGS: Turn text into vectors with Sentence Transformers, then measure meaning with cosine similarity and build semantic search.
🪐 4. RAG: Chat with your own PDFs. Learn chunking, embeddings, the Chroma vector store, grounded prompting and answers with source citations.
🪐 5. AGENT: Build AI agents with LangChain that use tools such as web search with Tavily and custom tools, and get real work done for you.
🪐 6. STRUCTURED: Get reliable structured output from LLMs, extracting clean data into Pydantic models with JSON schema.
🪐 7. MEMORY: Give your agents long-term memory with checkpointers and PostgreSQL, and keep separate conversation threads.
🪐 8. MIDDLEWARE: Control agent behavior with middleware, such as automatic conversation summarization.
🪐 9. LANGGRAPH: Design stateful workflows as graphs, such as a write → review → revise loop that improves until it's good enough.
🪐 10. MULTI-AGENT: Build a team of AI agents with the supervisor pattern, where one agent delegates to specialist agents.
📘 FLIGHT MANUAL (GitHub Repository)
All the course code is waiting for you on board:
https://github.com/mohamad-sw/ai-engineering-tutorial
⭐ If this journey helps you, please give the repo a star. It helps more passengers find this flight.
🧭 FLIGHT PLAN
🛫 BOARDING
00:00:00 – Course Introduction
00:04:27 – Setting Up the Development Environment
🪐 PLANET CHATBOT
00:13:35 – Building the Chat Function
00:18:10 – What is an LLM
00:26:51 – What is Hugging Face?
00:36:14 – Using the Groq Provider
00:41:43 – Using a Wrapper to Interact with LLMs
00:49:42 – Sending Requests to the LLM
01:04:31 – LLM Parameters
01:13:45 – Adding a Conversation Loop
01:18:26 – Conversation History
01:23:02 – Implementing Conversation History
01:26:20 – System Role
01:32:52 – What Are Tokens?
01:41:26 – Counting Total Tokens
01:48:01 – Making the Program Object-Oriented
01:57:45 – Calculating Cost
02:03:56 – Tiktoken
02:10:45 – Using Any Model's Tokenizer
02:19:49 – Streaming
02:30:31 – Context Window Management: The Sliding Window Strategy
02:44:06 – Summarizing Conversation History
🪐 PLANET CLASSIFIER
02:55:53 – Sentiment Analysis
03:03:39 – Prompt Engineering
03:10:23 – Natural Language Inference (NLI)
03:18:52 – BERT Model
03:25:10 – Using Pipeline to Run a Model
🪐 PLANET EMBEDDINGS
03:32:48 – What Are Embeddings?
03:40:43 – All-MiniLM-L6-v2
03:42:40 – Using Sentence Transformers Models
03:47:02 – Semantic Search
🪐 PLANET RAG
03:58:35 – Retrieval-Augmented Generation (RAG)
04:08:10 – The Retrieval Component
04:16:27 – Preparing the Knowledge Base
04:27:46 – Data Chunking
04:38:30 – Converting Chunks into Embeddings
04:46:09 – Using the Chroma Vector Database
04:59:29 – Generating IDs for Chunks
05:04:21 – Retrieving Chunks from the Vector Database
05:13:34 – Generation
🪐 PLANET AGENT
05:26:02 – What Is an AI Agent?
05:35:59 – Using the LangChain Wrapper
05:42:04 – Adding a Conversation Loop
05:48:24 – Building an Agent with create_agent Function
06:00:28 – Three Approaches to Building an Agent
06:04:37 – Building a Web Search Agent
06:11:20 – The MCP Protocol
🪐 PLANET STRUCTURED
06:18:43 – Dynamic Prompts
06:21:52 – The PromptTemplate Class
06:26:28 – The ChatPromptTemplate Class
06:35:15 – The MessagesPlaceholder Class
06:39:55 – Generating Structured Output
06:50:36 – Extracting Information with Structured Output
🪐 PLANET MEMORY
06:55:55 – Checkpointers
07:00:27 – Using InMemorySaver
07:06:28 – Switching Between Conversations
07:10:24 – Using PostgresSaver
07:18:18 – Reading Messages from the Database
🪐 PLANET MIDDLEWARE
07:22:32 – What Are Middlewares?
07:30:25 – Summarization Middleware
🪐 PLANET LANGGRAPH
07:35:03 – What Is LangGraph?
07:48:15 – Conditional Edges
07:55:42 – Automated Article Writing and Revision System
🪐 PLANET MULTI-AGENT
08:16:00 – Multi-Agent Systems
08:21:31 – The Supervisor Approach
🎟️ Boarding is free. If you know someone who dreams of becoming an AI Engineer, share this flight with them. Your captain will be grateful. 🙏
💬 Leave a message for the captain in the comments. He reads them all.
ℹ️ This course is not affiliated with or sponsored by any of the services or tools used in it.
Please fasten your seatbelts and keep your engines on. 🚀
#AIEngineering #LLM #RAG #AIAgents #LangChain #LangGraph #Python Close description




