⛓️ Exploring LangChain Runnables (LCEL)

Hey there! Welcome to my exploration of LangChain Runnables and the LangChain Expression Language (LCEL). After working with the older Chains interface, I wanted to dive into this newer, more flexible way of building AI workflows. The "pipe" (|) syntax isn't just for looks; it makes building complex, production-ready applications much more intuitive.

This repository is basically my personal journey and a collection of scripts where I break down how to use the most important Runnable components to create everything from simple sequences to a full-blown Retrieval-Augmented Generation (RAG) system.

---

🤔 Why Runnables? A Quick Note

Before diving in, I wanted to share why I focused on this. The LCEL and Runnable protocol are the future of LangChain. They offer huge advantages:

• Streaming Support: You can stream outputs token-by-token, which is amazing for creating responsive chat UIs.
• Asynchronous Operations: Easily run chains with async and await for better performance.
• Parallel Execution: Run parts of your workflow at the same time, which I've explored in one of the scripts.
• Automatic Tracing: LangChain automatically tracks every step of your Runnable sequence, which is a lifesaver for debugging.

---

✨ Core Concepts I've Explored

Here’s a breakdown of the key Runnable components I've implemented in this repository:

1. RunnableSequence (runnable_sequence.py):

   • This is the most fundamental concept. I started here to understand how to pipe components together: Prompt ➔ Model ➔ Parser. It's the modern way to do what SimpleChain used to do.

2. RunnablePassthrough (runnable_passthrough.py):

   • This one was a real "aha!" moment. It lets you pass inputs through a chain unchanged, which is crucial for RAG. I built a simple RAG chain here that takes a question, fetches relevant documents, and then "passes through" both the original question and the retrieved documents to the next step.

3. RunnableParallel (runnable_parallel.py):

   • Just like with the older chains, this lets you run multiple operations at the same time and get a combined dictionary output. I used it to create a chain that takes a topic and simultaneously generates both a question and a joke about it.

4. RunnableBranch (runnable_branch.py):

   • This is how you add conditional logic or "if/else" statements to your workflows. I created a chain that first classifies a user's question as either "Math" or "General Knowledge" and then routes it to a specialized chain for the correct answer.

5. Putting It All Together: A Full RAG Chain (retrievalQAchain.py):

   • This script is the final boss of this repository. It combines everything I learned to build a complete Retrieval-Augmented Generation (RAG) system. It takes a user's question, finds relevant information from a text file, and uses that context to generate an informed answer.

---

🛠️ Tech Stack

• Core Framework: LangChain & LangChain Expression Language (LCEL)
• LLM Provider: OpenAI
• Vector Store: FAISS (for in-memory semantic search in the RAG chain)
• Core Libraries: langchain-core, langchain-openai, python-dotenv

---

⚙️ Setup and Installation

1. Clone the repository:

   git clone [https://github.com/jsonusuman351/Langchain_Runnables.git](https://github.com/jsonusuman351/Langchain_Runnables.git)
   cd Langchain_Runnables

2. Create and activate a virtual environment:

   # It is recommended to use Python 3.10 or higher
   python -m venv venv
   .\venv\Scripts\activate

3. Install the required packages:

   pip install -r requirements.txt

4. Set Up Environment Variables:

   • Create a file named .env in the root directory.
   • Add your OpenAI API key to this file:

     OPENAI_API_KEY="your-openai-api-key"

---

🚀 How to Run the Scripts

Each script is a standalone example. Feel free to run them and see how each Runnable works.

• Basic Chaining:

  python runnable_sequence.py

• Chaining with Passthrough (for RAG context):

  python runnable_passthrough.py

• Parallel Execution:

  python runnable_parallel.py

• Conditional Logic:

  python runnable_branch.py

• RAG System: Make sure you have a PDF file in your project directory for this to work.

  python retrievalQAchain.py

---

🔬 A Tour of My Runnable Experiments

I've organized the scripts based on the core Runnable concept they demonstrate.

Click to view the code layout

Langchain_Runnables/
│
├── runnable_sequence.py       # The basic "pipe" workflow
├── runnable_passthrough.py    # Passing data through a chain
├── runnable_parallel.py       # Running operations at the same time
├── runnable_branch.py         # Adding if/else logic to chains
├── retrievalQAchain.py        # A complete RAG implementation
│
├── pdf_reader.py              # Helper script for loading PDFs
├── docs.txt                   # Sample text file for RAG
│
├── requirements.txt
├── .env                       # (You need to create this for your API key)
└── README.md

---

---