IBM Tutorials Repository

Welcome to the IBM Tutorials repository - your comprehensive resource for learning cutting-edge AI, machine learning, and generative AI technologies through hands-on tutorials and projects.

🎯 What You'll Find Here

This repository contains 60+ tutorials organized by learning intent, covering:

• Retrieval-Augmented Generation (RAG) - Build intelligent document Q&A systems
• AI Agents & Orchestration - Create autonomous agents with LangChain, LangGraph, and more
• Multi-Agent Systems - Implement collaborative AI systems with CrewAI, BeeAI, and AutoGen
• Prompt Engineering - Master the art of effective LLM communication
• Multimodal AI - Work with vision, speech, and multimodal models
• Tool & Function Calling - Extend LLM capabilities with external tools
• Guardrails & Safety - Build responsible AI systems with security and safety mechanisms
• Time Series & Forecasting - Apply AI to temporal data
• Text Processing & NLP - Classic and modern NLP techniques
• Full-Stack AI Applications - Complete end-to-end AI projects
• Observability & Monitoring - Track and optimize AI system performance
• Data Streaming & Event Processing - Build real-time data pipelines with Kafka and Confluent
• IBM Bob - Master the AI-powered coding assistant for documentation and automation

🚀 Quick Start

Prerequisites

• Python 3.10 - 3.13 (Python 3.11 recommended)
• IBM watsonx.ai account (for most tutorials) - Sign up here
• Git for cloning the repository

Installation

1. Clone the repository:

   git clone https://github.com/IBM/ibmdotcom-tutorials.git
   cd ibmdotcom-tutorials
   

2. Navigate to a tutorial:

   cd tutorials/01-rag-and-retrieval  # or any other category
   

3. Follow the tutorial's setup instructions:

   • Each tutorial includes its own setup and dependency installation instructions
   • Most tutorials require IBM watsonx.ai credentials
   • See individual tutorial READMEs for specific requirements

4. Start learning:

   • Open any .ipynb file in your IDE (VS Code, PyCharm, etc.)
   • Follow the tutorial's step-by-step instructions

📚 Tutorial Categories

01 - RAG and Retrieval

Build intelligent systems that answer questions from your documents using vector search and embeddings.

Featured Tutorials:

• LangChain RAG - Your first RAG system
• Agentic RAG - RAG with reasoning capabilities
• RAG Evaluation with Ragas - Measure performance

02 - Agents & Orchestration

Create autonomous AI agents that can plan, reason, and execute complex tasks.

Featured Tutorials:

• LLM Agent Orchestration
• Building Agentic Workflows with LangGraph
• Text Classification Agent with watsonx Orchestrate - Build sentiment analysis agents
• SQL Agent Application - Full-stack agent app

03 - Multi-Agent Systems

Implement collaborative AI systems where multiple agents work together.

Featured Projects:

• CrewAI Retail Example
• BeeAI Agent-to-Agent Tutorial
• Customer Service Call Analysis
• ChatDev with watsonx

04 - Prompt Engineering

Master techniques for effective LLM communication and optimization.

Featured Tutorials:

• Role Prompting - Assign roles for better responses
• Active Prompting - Improve accuracy on uncertain cases with watsonx Orchestrate
• Prompt Chaining with LangChain - Chain prompts for complex tasks

05 - Multimodal AI

Work with vision, speech, and multimodal models for diverse AI applications.

Featured Tutorials:

• Granite Vision PPT Analyzer
• AI Personal Trainer with Llama
• Granite Speech 3.3

06 - Tool & Function Calling

Extend LLM capabilities by integrating external tools and APIs.

07 - Guardrails & Safety

Build responsible AI systems with safety mechanisms and content filtering.

Featured Tutorials:

• AI Agent Security - Secure AI agents with authentication, RBAC, and data protection
• LLM Guardrails - Input/output filtering and content moderation
• Granite Guardian Web App - Real-time content filtering application

08 - Time Series & Forecasting

Apply AI to temporal data for forecasting and analysis.

09 - Text Processing & NLP

Classic and modern natural language processing techniques.

10 - Machine Learning Foundations

Core ML concepts and techniques.

11 - Model Context Protocol

Work with MCP servers and IBM Bob integration.

12 - Observability & Monitoring

Track, monitor, and optimize AI system performance.

Featured Tutorials:

• watsonx Orchestrate with AgentOps
• watsonx Observability with Langfuse

13 - Full-Stack Applications

Complete end-to-end AI applications and projects.

Featured Projects:

• AI Stylist
• TTRPG AI
• Silly Story Time

14 - LoRA & Fine-Tuning

Customize models for your specific use cases.

15 - Docling

Parse, convert, and process documents using IBM's open-source Docling toolkit.

Featured Tutorials:

• Convert Unstructured Data to Structured Data - Transform scanned documents into structured formats
• Docling Granite Question Answering - Document Q&A with Granite 3.1
• DeepSeek RAG Reasoning with Docling - Advanced RAG with Docling

16 - IBM Bob

Master IBM Bob, the AI-powered coding assistant for documentation, development, and automation.

Featured Tutorials:

• AI Documentation with IBM Bob - Automatically generate and maintain project documentation
• MCP Server Integration - Build and integrate Model Context Protocol servers

17 - Data Streaming

Build real-time event streaming platforms and data pipelines using Apache Kafka and Confluent Cloud.

Featured Tutorials:

• Confluent Modernization - Migrate from on-premises Kafka to Confluent Cloud with real-time fraud detection

🎓 Learning Paths

Beginner Path: Getting Started with AI

Perfect for those new to AI and LLMs. Start here to build foundational knowledge.

1. LangChain RAG - Build your first document Q&A system
2. Docling Granite Question Answering - Document processing with Granite 3.1
3. LLM Agent Orchestration - Create autonomous AI agents
4. Function Calling - Extend LLM capabilities with tools
5. Role Prompting - Master effective prompt techniques

Intermediate Path: Building AI Systems

For developers ready to build more sophisticated AI applications.

1. Agentic RAG - RAG with reasoning capabilities
2. Building Agentic Workflows with LangGraph - Advanced agent orchestration
3. Text Classification Agent with watsonx Orchestrate - Build sentiment analysis agents
4. RAG Evaluation with Ragas - Measure and optimize RAG performance
5. Multimodal AI with Granite Vision - Work with vision models
6. LLM Guardrails - Build responsible AI systems
7. Convert Unstructured Data with Docling - Transform documents into structured formats

Advanced Path: Production AI

Master enterprise-grade AI systems with multi-agent collaboration and observability.

1. Multi-Agent Customer Service Analysis - Collaborative AI systems
2. BeeAI Agent-to-Agent Communication - Agent collaboration protocols
3. AI Agent Security - Secure agents with authentication and RBAC
4. watsonx Orchestrate with AgentOps - Monitor agent performance
5. watsonx Observability with Langfuse - Track and optimize AI systems
6. SQL Agent Application - Full-stack agent application
7. DeepSeek RAG Reasoning with Docling - Advanced RAG techniques

Specialized Path: Multimodal & Full-Stack AI

Build complete applications with vision, speech, and multimodal capabilities.

1. Granite Speech 3.3 - Speech processing and transcription
2. AI Personal Trainer with Llama - Vision-based fitness analysis
3. AI Stylist - Complete multimodal fashion advisor
4. Silly Story Time - Interactive storytelling application
5. Granite Guardian Web App - Real-time content filtering

IBM Bob Path: AI-Powered Development

Master IBM Bob for documentation, automation, and development workflows.

1. AI Documentation with IBM Bob - Automatically generate project documentation
2. MCP Server Integration - Build Model Context Protocol servers
3. Model Context Protocol Basics - Understand MCP fundamentals

🛠️ Technologies Used

• IBM watsonx.ai - Enterprise AI platform
• IBM Granite Models - Open-source foundation models
• LangChain - LLM application framework
• LangGraph - Agent workflow orchestration
• LlamaIndex - Data framework for LLMs
• CrewAI - Multi-agent orchestration
• BeeAI - Agent framework
• AutoGen - Multi-agent conversations
• Ollama - Local LLM deployment
• Chroma, Milvus - Vector databases
• Ragas - RAG evaluation framework

🤝 Contributing

We welcome contributions! Whether you want to:

• 🐛 Report a bug
• 💡 Suggest a new tutorial
• 📝 Improve documentation
• 🔧 Submit a pull request

Please see our Contributing Guide for detailed instructions on how to contribute, including setup, development workflow, and code quality standards.

Also see our Code of Conduct for community guidelines.

📊 Repository Structure

ibmdotcom-tutorials/
├── tutorials/              # All tutorials organized by category
│   ├── 01-rag-and-retrieval/
│   ├── 02-agents-and-orchestration/
│   ├── 03-multi-agent-systems/
│   ├── 04-prompt-engineering/
│   ├── 05-multimodal-ai/
│   ├── 06-tool-calling-and-function-calling/
│   ├── 07-guardrails-and-safety/
│   ├── 08-time-series-and-forecasting/
│   ├── 09-text-processing-and-nlp/
│   ├── 10-machine-learning-foundations/
│   ├── 11-model-context-protocol/
│   ├── 12-observability-and-monitoring/
│   ├── 13-full-stack-applications/
│   ├── 14-lora-and-fine-tuning/
│   ├── 15-docling/
│   ├── 16-ibm-bob/
│   ├── 17-data-streaming/
│   └── shared-assets/      # Shared data, images, and resources
├── .github/                # GitHub workflows and assets
└── README.md              # This file

🔗 Useful Links

• IBM watsonx - AI platform
• IBM Granite Models - Open-source LLMs
• IBM Think - Technical articles
• GitHub Discussions - Community Q&A

📝 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

🙋 Support

• Questions? Open a GitHub Discussion
• Issues? Report them in GitHub Issues

🌟 Star History

If you find these tutorials helpful, please consider giving us a star! ⭐

---

Maintained by: IBM.com Technical Content Team
Last Updated: February 2026