AgenticAI: An Experimental Approach to Autonomous Intelligence

A comprehensive collection of AI agent implementations demonstrating various LangChain and LangGraph capabilities for building intelligent, multi-step systems.

🚀 Quick Start

Prerequisites

• Python 3.8+ installed
• Basic command line knowledge
• Recommended IDE: Visual Studio Code

Clone Repository & Setup Environment

1. Clone the Git Repository

# Clone the repository
git clone https://github.com/kazhian/agentic_ai.git

# Navigate into the project directory
cd agentic_ai

2. Create Virtual Environment

# Create virtual environment (recommended)
python -m venv .venv

# Activate virtual environment
# On Windows:
.venv\Scripts\activate
# On macOS/Linux:
source .venv/bin/activate

# Verify activation (you should see (.venv) in your prompt)

Why Virtual Environment?

• Isolates project dependencies
• Prevents conflicts with system packages
• Easy to recreate and share
• Best practice for Python development

IDE Setup (Recommended)

Visual Studio Code Setup:

# Install VS Code (if not already installed)
# Visit: https://code.visualstudio.com/

# Install essential VS Code extensions:
# 1. Python Extension (Microsoft)
# 2. Jupyter Extension (Microsoft)
# 3. Python Docstring Generator
# 4. GitLens (for better code navigation)

Why VS Code?

• Excellent Python support with IntelliSense
• Built-in Jupyter notebook support
• Integrated terminal for running commands
• Great debugging capabilities
• Free and widely used

Setup Instructions

3. Install Python Dependencies

# Install required packages (make sure .venv is activated)
pip install langchain langchain-openai langgraph langchain-community python-dotenv tavily-python crewai

4. Get API Keys

OpenAI API Key:

• Visit OpenAI Platform
• Sign up/login and create a new API key
• Copy the key (starts with sk-proj-)

Tavily API Key:

• Visit Tavily AI
• Sign up and get your API key
• Copy the key (starts with tvly-)

LangSmith API Key:

• Visit LangSmith or LangChain LangSmith
• Sign up/login and create a new API key
• Copy the key

5. Configure Environment

# Copy the template file
cp template.env .env

# Edit the .env file with your actual API keys
# OPENAI_API_KEY='sk-proj-your-actual-key-here'
# TAVILY_API_KEY='tvly-your-actual-key-here'
# LANGSMITH_API_KEY='your-langsmith-api-key-here'

6. Verify Setup

# Test your configuration
python 01_setup/01_direct_interaction.py

If the setup is correct, you'll see a successful AI response. If not, check your API keys and Python installation.

7. Setup Markdown Display (Optional)

For better markdown rendering when viewing CrewAI outputs:

VS Code Built-in Preview (Recommended):

• Open any .md file in VS Code
• Press Ctrl+Shift+V (Windows/Linux) or Cmd+Shift+V (Mac)
• Or click the preview icon in the top right

Python Libraries for Enhanced Display:

# For terminal rendering
pip install rich

# For HTML conversion
pip install markdown

Desktop Apps:

• Typora (paid, beautiful)
• Mark Text (free)
• Obsidian (free, powerful)

Online Viewers:

• https://markdownlivepreview.com/
• https://dillinger.io/
• GitHub Gists

📁 Project Structure

├── 01_setup/              # Basic setup and direct API interaction
├── 02_langchain/          # LangChain fundamentals and chains
├── 03_lc_tool_integ/      # Tool integration examples
├── 04_lc_memory/          # Memory and conversation persistence
├── 05_multi_step_agent/   # Multi-step agent workflows
├── 06_crewai_multi_agent/ # CrewAI multi-agent orchestration
├── 07_autonomous_agents/  # AutoGPT-style autonomous agents
├── 08_agent_evaluation/   # Evaluation and LangSmith tracing
├── 09_industry_case/      # Real-world industry applications
└── 11_ethical_consideration/ # AI ethics and safety

🧭 Course Module Map

This course is organized around the topics for AgenticAI: An Experimental Approach to Autonomous Intelligence.

1. Setting up Agentic AI Development Environment
   • Covers virtual environments, dependency installation, API key configuration, and direct model interaction.
2. Building Agents with LangChain: Simple Question Answering
   • Includes basic LangChain agent patterns and Q&A examples.
3. Tool Integration with LangChain Agents: Web Search
   • Demonstrates how to add external tool capabilities like web search.
4. Memory Management in LangChain Agents: Conversational History
   • Teaches agent memory, chat history, and session persistence.
5. Creating Multi-Step Agents with LangChain Chains
   • Shows how to build multi-step workflows and chain-based agents.
6. Orchestrating Multi-Agent Systems with CrewAI
   • Walks through coordinated multi-agent workflows using CrewAI.
7. Exploring Autonomous Agents with AutoGPT
   • Introduces autonomous agent concepts and AutoGPT-style execution.
8. Evaluating Agent Performance using LangSmith (or similar tools)
   • Covers tracing and evaluation concepts using notebooks and examples.
9. Building an Agent for a Specific Industry Use Case — Customer Support
   • Demonstrates a customer support agent workflow and practical application.
10. Building an Agent for a Specific Industry Use Case — Financial Data Analysis

• Describes a potential extension for building a domain-specific financial agent.

11. Exploring Ethical Considerations in Agent Behavior

• Covers accountability, transparency, bias, privacy, and safety.

12. Final Project: Developing an Agentic AI Application for a Real-World Problem

• The capstone project concept is to combine these modules into a complete, real-world agentic application.

Note: Module 08_agent_evaluation/ has been added to provide dedicated LangSmith tracing and agent evaluation content.

🎯 Learning Path

1. Start Here: 01_setup/ - Verify your environment works
2. Basics: 02_langchain/ - Learn LangChain fundamentals
3. Tools: 03_lc_tool_integ/ - Add external capabilities
4. Memory: 04_lc_memory/ - Enable conversation persistence
5. Workflows: 05_multi_step_agent/ - Build multi-step agents
6. Multi-Agent: 06_crewai_multi_agent/ - Orchestrate agent crews
7. Autonomous: 07_autonomous_agents/ - Explore AutoGPT-style autonomous agents
8. Applications: 09_industry_case/ - Real-world implementations
9. Ethics: 11_ethical_consideration/ - Responsible AI development

🛠️ Key Technologies

• LangChain: Framework for building AI applications
• LangGraph: Stateful workflow orchestration
• CrewAI: Multi-agent collaboration and orchestration
• OpenAI: GPT models for language understanding
• Tavily: Web search and information retrieval
• Python: Core programming language

📚 What You'll Learn

• Building conversational AI agents
• Creating multi-step workflows
• Orchestrating multi-agent systems
• Integrating external tools and APIs
• Managing conversation memory
• Implementing industry-specific solutions
• Understanding AI ethics and safety

🔧 Common Issues

API Key Problems:

• Ensure keys are copied correctly without extra spaces
• Check that your OpenAI account has credits
• Verify Tavily API key is valid

Python Issues:

• Use Python 3.8 or higher
• Install packages in a virtual environment (recommended)
• Update pip: pip install --upgrade pip

Virtual Environment Issues:

• Make sure you activated the virtual environment before installing packages
• If activation fails, try: python -m venv .venv --clear then reactivate
• On Windows, use PowerShell or Command Prompt as Administrator

VS Code Issues:

• Restart VS Code after installing Python extension
• Select correct Python interpreter: Ctrl+Shift+P → "Python: Select Interpreter"
• For Jupyter notebooks, ensure Jupyter extension is installed

Import Errors:

• Make sure you're in the project root directory
• Check that all packages installed successfully
• Restart your terminal after installation

🎯 Get Started Now

# 1. Clone the repository
git clone https://github.com/kazhian/agentic_ai.git
cd agentic_ai

# 2. Create and activate virtual environment
python -m venv .venv
# On Windows: .venv\Scripts\activate
# On macOS/Linux: source .venv/bin/activate

# 3. Install VS Code: https://code.visualstudio.com/
# 4. Install VS Code extensions: Python, Jupyter, Python Docstring Generator, GitLens

# 5. Install dependencies (in virtual environment)
pip install langchain langchain-openai langgraph langchain-community python-dotenv tavily-python crewai

# 6. Set up your API keys in .env file
cp template.env .env
# Edit .env with your actual keys

# 7. Test everything works
python 01_setup/01_direct_interaction.py

# 8. Setup markdown display (optional)
# Follow the "Setup Markdown Display" section above

📖 Additional Resources

• LangChain Documentation
• LangGraph Guide
• OpenAI API Docs
• Tavily API Guide

---

🎓 Learning Goal: Master the art of building intelligent AI agents that can reason, remember, and act autonomously in real-world scenarios.

Start with the setup verification, then progress through the numbered directories to build your skills progressively!