LangChain Practice Hub 🦜🔗

This repository serves as a collection of mini-projects designed to master LangChain concepts, ranging from basic prompts to advanced chains and agents. Each module focuses on specific real-world applications and "Clean Code" practices.

📂 Active Modules

Project	Description	Documentation
Smart Assistant	High-level text analysis tool (Email, Tech Explainer, Summary) using Gemini Pro.	Read Docs
Conversation AI	Stateful agent with memory and tools (Order Search, User Info) powered by ReAct logic.	Read Docs

---

⚡ Quick Start

This project uses uv for blazing fast dependency management. We do not use pip or poetry.

1. Pre-requisites

Ensure you have uv installed:

# On macOS/Linux
curl -LsSf https://astral.sh/uv/install.sh | sh

# On Windows
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

2. Installation

Clone the repo and sync dependencies:

# Install dependencies & create virtualenv automatically
uv sync

3. Environment Setup 🔐

You need to configure your API keys before running any project.

1. Copy the example configuration:

   cp .env.example .env

2. Open .env and fill in the required variables:

   # .env
   GOOGLE_API_KEY=AIzaSy...     # Get this from Google AI Studio
   GOOGLE_API_MODEL=gemini-pro  # Model version
   TEMPERATURE=0.7              # Creativity (0.0 - 1.0)

4. Running Projects 🚀

Run any script using uv run. It automatically handles environment activation.

# Run the main application
uv run app.py

---

📚 Educational Purpose

The code within this repository is structured for educational clarity:

• Separation of Concerns: Config, Core, Services are isolated.
• Modern Python: Uses Pydantic, Type Hinting, and LCEL.
• Production Ready: Concepts can be adapted for real-world deployment.

📄 License

This project is open-source and available under the terms of the MIT License.