Generative AI Engineer: LangChain, AI Agents & 12 Projects

A beginner AI bootcamp in Python: become a GenAI engineer and AI developer by building AI agents, RAG and more.

  • 0.0 (0 Reviews)
  • 0 Students
  • 127 Lessons
  • Ardit Sulce | 600,000+ Students

Course Overview

This AI bootcamp teaches you how to build generative AI apps and AI agents with Python. You will use LangChain 1, the newest version of LangChain, together with models from Google Gemini, OpenAI and Ollama.

The course is built around 12 projects. Each project is a working program that you write step by step together with me. Before each project, I explain the idea behind it. For example, what an AI agent is, how an agent remembers a conversation, or how a RAG system finds the right part of a document. Then we write the code.

The 12 projects

  1. A weather AI agent that gets real weather data for the user's location

  2. An AI agent web app built with Flask, with chat history

  3. Your own LLM API service that runs a free local model with Ollama

  4. A daily AI report of new spreadsheet rows, sent to you by email

  5. A Python program that reads and edits private Google Sheets

  6. A recipe generator web app that reads a photo of ingredients

  7. An AI image generator with Google Gemini models

  8. A voice assistant that listens to you and answers out loud

  9. A chat app that answers questions about your own PDF, text and Markdown files

  10. A full RAG system with embeddings and a vector store

  11. A newsletter writer that searches the web with Tavily

  12. A Claude Code clone: a coding agent that reads and writes files on your computer

Between the projects, you learn the skills they need. These include tools and system prompts for agents, saving agent memory in SQLite and Supabase, structured JSON output with Pydantic, and running free models on your own computer with Ollama.

Who this course is for

This is a beginner course in AI. You need basic Python: variables, loops, functions and lists. You do not need any experience with AI or LLMs. We start from how LLMs work and build up from there. It is a good first step if you want to become an AI developer or move into AI engineering.

Practice in PyCharm

The course comes with a free and optional practice course on JetBrains Academy. There you solve coding tasks inside the PyCharm IDE as you follow the videos.

About API costs

Most projects run on free plans: the Gemini API, Tavily, Supabase and OpenWeather. A few lectures use paid models: image generation with Google, the voice assistant with OpenAI, and one image example with OpenAI. The cost is small and you pay only for what you use. You can also just watch those lectures. The Claude Code clone uses OpenAI by default, and I show you how to switch it to Gemini.

By the end of the course, you will have 12 working AI programs and the skills to build your own. I look forward to seeing you in the first lecture.

See more

FAQ

FAQ area empty

Course curriculum

20 chương 127 bài học 11 giờ 34 phút 7 bài học thử miễn phí
Chương 1 Start Here: LLMs, AI Apps, RAG and AI Agents Explained
Chương 2 AI Agents with LangChain 1: Tools and System Prompts
  • What We Will Build: A Weather AI Agent 1:00
  • What Is an AI Agent? 2:21
  • Building an AI Agent with LangChain 9:12
  • Adding Tools to the AI Agent 8:02
  • Creating an Agent With Multiple Tools 6:24
  • How to Write a Good Agent System Prompt 6:32
  • Practice: Tool Calling and Agents
Chương 3 Project 1: Weather AI Agent with Real Data and User Location
  • What We Will Build: Real Weather for the User's Location 2:26
  • Getting Real Weather Data from a Weather API 9:27
  • Agent Getting the User's Location 5:10
  • Improving the Agent Output 8:35
  • Practice: Prompt Engineering
Chương 4 Giving the AI Agent Memory (Chat History)
  • What We Will Build: An Agent That Remembers 1:31
  • Enabling Agent Memory 7:53
  • Understanding how Memory Works in Agents 8:11
  • AI Agents with Multiple Conversations 5:50
  • Continuous Agent Conversation with Memory 4:33
Chương 5 Saving Agent Memory in SQLite and Supabase (PostgreSQL)
  • Why Agents Need Persistent Memory 1:01
  • Saving and Retrieving Chat History in SQLite 11:33
  • Architecture Using a PostgreSQL Database 3:08
  • Setting up a PostgreSQL Database on Supabase 5:45
  • Saving the Conversation in Supabase 4:37
Chương 6 Project 2: AI Agent Web App with Flask
  • What We Will Build: The AI Agent as a Web App 1:38
  • From CLI to Web App 2:15
  • Creating a Flask Web App 3:56
  • Building the AI Agent Homepage 6:14
  • Receiving User Messages with a POST Request 7:16
  • Getting the Agent Response from the Web App 7:25
  • Showing the Conversation with a Flask Session 12:31
  • Styling the Chat Page with CSS 3:24
  • Making the Agent Get the User's GPS Location 12:32
  • Practice: Guardrails and AI Security
Chương 7 Free and Private Local LLMs with Ollama
  • Why Local Models: Free, Private and Under Your Control 1:01
  • Remote vs Local LLMs and Which Models to Download 5:15
  • Installing and Using Ollama 7:59
  • Downloading More Models with the Ollama Command Line 6:47
  • Using the Local Model Through Python and LangChain 6:53
Chương 8 Project 3: Your Own LLM API Service with Flask and Ollama
  • What We Will Build: Your Own LLM API 2:21
  • How an LLM API Works 4:47
  • Preparing the Environment 2:11
  • Building the API with Flask and Ollama 11:45
  • LLM API with Chat History Support 11:17
  • Practice: Shipping AI to Production
Chương 9 Structured JSON Output from LLMs and Agents (Pydantic)
  • Why Do We Need Structured Data? 3:02
  • How Structured Output Works (Pydantic Models) 1:53
  • Returning Structured Output with LangChain 10:55
  • Returning Structured Output from Agents 10:17
  • Practice: Structured Output and Extraction
Chương 10 Project 4: Daily AI Spreadsheet Report Sent by Email
  • What We Will Build: A Daily AI Spreadsheet Report 1:40
  • Setting Up a Spreadsheet and Google Credentials 5:58
  • Load Spreadsheet Data into Python 7:52
  • Detecting Newly Added Rows 10:28
  • Summarizing the New Rows with an LLM 14:28
  • Creating a Gmail App Password 2:12
  • Sending the AI Summary by Email 6:32
  • Scheduling the Task for Automatic Execution 5:10
  • Practice: Workflows, Multi-Agent and MCP
Chương 11 Project 5: Read and Edit Private Google Sheets with Python
  • What We Will Build: View, Add and Edit Sheet Rows 1:15
  • Connecting Google Sheets with Python 4:46
  • Reading Data from a Private Google Sheet 11:52
  • Adding New Rows to the Spreadsheet 7:37
  • Updating Existing Rows and Cells 6:17
Chương 12 Reading Images with Gemini and OpenAI Vision Models
  • What You Will Learn: AI That Understands Images 1:16
  • Sending Text to the Gemini Model 3:13
  • Interpreting Images with Google Gemini Models 8:33
  • Interpreting Images with OpenAI Models 6:22
Chương 13 Project 6: Recipe Generator Web App from Food Photos (Gradio)
  • What We Will Build: Recipes from a Photo of Ingredients 1:14
  • Generating Recipes from Ingredients in Image 6:32
  • Gradio Basics: A Simple Web App 6:43
  • Creating the Recipe Generator Web App 6:55
Chương 14 Project 7: AI Image Generator with Python and Google Gemini
  • What We Will Build: Images Generated with Python 1:06
  • A Tour of the OpenAI Platform: Models and Prices 4:22
  • A Tour of Google AI Studio 2:58
  • Understanding Google AI Model Prices 6:43
  • Setting Up Billing in Google AI Studio 2:31
  • Generating Images with Gemini in Python 8:19
Chương 15 Project 8: AI Voice Assistant You Can Talk To
  • What We Will Build: A Voice Assistant You Talk To 1:40
  • How the Voice Assistant Works, Step by Step 2:08
  • Record Audio with Python 10:33
  • Convert Speech to Text with OpenAI 5:52
  • Sending the Converted Text to the LLM 4:41
  • Text to Speech Conversion with OpenAI 6:21
Chương 16 Project 9: Chat with Your Own Documents (Context Injection)
  • What We Will Build: Chat with Your Documents 4:03
  • The Architecture of the System 3:24
  • Initializing the LLM 3:52
  • System Prompt and Context Injection 9:00
  • Loading PDF, Text and Markdown Files into Python 16:46
  • Creating the Context 7:00
  • Getting Answers from the LLM About the Documents 3:28
  • Asking Follow-Up Questions About the Documents 4:43
Chương 17 Project 10: Full RAG System with Embeddings and a Vector Store
  • How a RAG System Works 3:30
  • Setting Up the LLM 5:12
  • Setting up Embeddings and Vector Store 7:14
  • Creating Embeddings from Documents 9:13
  • Getting Answers from the RAG 7:07
  • Practice: RAG Systems
Chương 18 Project 11: AI Newsletter Writer with Web Search (Tavily)
  • What We Will Build: A Newsletter Writer That Searches the Web 2:00
  • Getting the Tavily and Gemini API Keys 2:53
  • What is Tavily 3:25
  • Searching the Web from Python with Tavily 11:19
  • Performing Multiple Searches 7:10
  • Writing a Newsletter from the Search Results 9:34
  • Practice: Evals and Error Analysis
Chương 19 Project 12: Build a Claude Code Clone (Local Coding Agent)
  • What We Will Build: A Claude Code Clone in Python 4:53
  • The ReAct Pattern: How Agents Think and Act 4:42
  • Creating a Simple Agent 8:38
  • Adding File Access Capability 11:06
  • Streaming Agent Updates 7:46
  • Streaming Updates on Tool Use 7:28
  • Streaming Updates on Tool Call Results 5:27
  • Displaying the Final Agent Answer 6:03
Chương 20 Project 12, Part 2: Finishing the Claude Code Clone
  • Recap: The Coding Agent So Far 6:01
  • Styling the Console Output with Rich 4:33
  • Styling the Agent's Final Answer 5:32
  • Letting the Agent Write Files 6:24
  • Letting the Agent Read Files and Create Folders 7:25
  • Finalizing the Agent 5:32
  • Your Next Step: The AI Engineer Path

Requirment

  • Basic Python Knowledge: You should understand variables, loops, functions, and lists.

  • No AI Experience Required: We will cover the AI and LLM concepts from the ground up.

Outcomes

  • Build AI agents in Python with LangChain 1 that call your own Python functions as tools.

  • Give agents memory and save their conversations in SQLite and in a PostgreSQL database on Supabase.

  • Turn an AI agent into a Flask web app with chat history and the user's location.

  • Run free and private LLMs on your own computer with Ollama, and serve one through your own API.

  • Get structured JSON output from LLMs and agents with Pydantic models.

  • Automate Google Sheets with Python and send yourself a daily AI summary of new rows by email.

  • Build apps that read images, generate images and talk with the user by voice.

  • Chat with your own documents, then build a full RAG system with embeddings and a vector store.

  • Search the web from Python with Tavily and turn the results into a newsletter with Gemini.

  • Build a Claude Code clone in Python: a coding agent that reads and writes files on your computer.

Instructor

...
Ardit Sulce | 600,000+ Students

0.0

  • ... 0 Student
  • ... 1 Course
  • ... 0 review

View Details

Reviews

Rate this course :

Remove all