AI Agent Course - Build a Language‑Learning Agent with OpenAI, LangGraph, Ollama & MCP

Build an AI agent that learns alongside you

In this free, end‑to‑end course, you’ll build a powerful language‑learning AI agent from scratch using Python, LangGraph, OpenAI, Ollama, and MCP. Over this hands‑on tutorial, we’ll create a ReAct‑style agent that sources vocabulary, performs accurate translations, and automatically generates Anki flashcards.

You’ll learn how to:
• Build a ReAct agent using LangGraph
• Connect agents to external tools with MCP
• Combine proprietary and local LLMs (OpenAI & Ollama)
• Design real‑world AI agent workflows

No NLP background required — we explain both the how and the why. By the end of this course, you’ll go from an empty Python project to a fully functional AI assistant you can adapt for your own AI projects.

Download PyCharm for free, the only Python IDE you need to build data models and AI agents: https://jb.gg/ai-agents-course

The full code for this tutorial, as well as the resources used, can be found in this GitHub repo: https://github.com/t-redactyl/....language-learning-ag
You can connect with Jodie, as well as see more of her work, at http://t-redactyl.io

Resources:
https://github.com/eymenefealt....un/all-words-in-all-
https://www.kaggle.com/datasets
https://www.kaggle.com/datasets?tags=13204-NLP
https://archive.ics.uci.edu/datasets
https://huggingface.co/learn/agents-course/

Timestamps
00:00 - Course Intro: What We’ll Build, AI agent tech stack
00:43 - What You’ll Learn: Building an AI Language Learning Agent
01:41 – Who This AI Agent Tutorial Is For (Python Prerequisites)
02:42 – Instructor Introduction: NLP & Data Science Background
03:01 – Why Use PyCharm for AI Agents & Data Science

Environment & Project Setup
03:36 – Installing PyCharm and AI Assistant
05:45 – Creating a Python Project with Virtual Environments

Dataset Selection & Preparation
07:26 – Choosing a Multilingual Vocabulary Dataset
08:46 – Best NLP Datasets: Kaggle & UCI Repositories
10:02 – Cloning and Organizing NLP Datasets in PyCharm

Installing Python & AI Dependencies
13:06 – Installing NLP Libraries: pandas, SpaCy, wordfreq
15:09 – Installing LangChain, LangGraph & MCP Libraries

Data Exploration & Analysis
16:56 – Exploring NLP Data with Jupyter Notebooks
18:01 – Analyzing Vocabulary Size Across Languages
20:45 – Visualizing Word Counts with Pandas Charts
22:08 – Identifying Data Problems in Multilingual Word Lists
24:36 – Introduction to SpaCy for Natural Language Processing

Cleaning the Word Lists
29:02 – Inspecting and Debugging Raw Vocabulary Data
31:48 – Removing Noise: Basic Text Cleaning Techniques
33:06 – Lemmatizing Words with SpaCy Models
35:10 – Using Zipf’s Law Overview to Filter Rare Words
37:26 – Word Frequency Analysis with wordfreq and SpaCy
42:35 - Understand Word Frequencies with wordfreq

Final Dataset Creation
45:22 – Building a Complete NLP Cleaning Pipeline
49:01 – Validating Results with a Spanish Dataset
50:01 – Comparing Raw vs Cleaned NLP Data

ReAct Agent Basics
54:40 – From Clean Data to an AI Agent
55:45 – What Is an AI Agent? Core Concepts
56:19 – Thought-Action-Observation Loop Explained
58:11 – Types of AI Agents

Large Language Models (LLMs) Explained
1:00:56 – How Large Language Models Understand Language
1:01:01 – Word Embeddings & Word2Vec Explained
1:05:24 – Why Word Embeddings Fail Without Context
1:06:05 – Transformers & Self-Attention Explained
1:10:23 – GPT Models and Decoder-Only Architectures

Reasoning Models for AI Agents
1:11:52 – Why Reasoning Models Power AI Agents
1:12:08 – How Reasoning Models Are Trained (Chain-of-Thought)
1:17:04 – When Not to Use Reasoning Models

1:20:24 How to Build a ReAct Agent with LangGraph
1:21:31 Agent State, Memory & Tools Explained
1:23:34 Choosing Between GPT-4 and Open-Source Models
1:26:36 How to Manage OpenAI API Keys Securely

1:30:09 How to Build Custom Tools for LangGraph Agents
1:33:14 Auto-Generating Tool Docstrings with AI
1:35:01 Improving Agent Reliability with System Prompts

1:38:13 Building and Connecting a LangGraph Agent Graph
1:41:21 Running an AI Agent End-to-End
1:43:02 How to Debug AI Agents in PyCharm
1:44:37 Visualizing Agent Execution Graphs

1:47:36 How to Run AI Agents Locally with Ollama
1:49:24 Choosing the Best Open-Source Reasoning Model
1:51:04 Installing and Managing Ollama Models
1:53:19 GPT-4 vs Ollama: Model Comparison for Agents

1:58:12 Switching LangGraph Agents from OpenAI to Ollama
2:00:22 Testing a Fully Local AI Agent

2:11:31 Adding Difficulty-Aware Vocabulary Tools
2:14:01 Handling Ambiguous User Requests in AI Agents
2:19:56 Testing AI Agents with Natural Language Prompts

2:24:48 How to Translate Words Using an LLM Tool
2:27:34 Building a Translation Tool with Ollama
2:32:28 Parsing Structured Output from LLMs

2:37:02 Multi-Step Tool Use in ReAct Agents
2:39:34 Handling Errors and Non-Determinism in AI Agents

2:41:40 What Is MCP (Model Context Protocol)?
2:42:13 Connecting AI Agents to External Tools with MCP