🍽️ LangChain Groq Restaurant Agent

AI that names your restaurant, builds your menu, and thinks for itself

---

🧠 What Is This?

This project combines two things - a working AI web app and a deep dive into LangChain agents.

Pick a cuisine. The AI names your restaurant, then builds a full menu for it powered by LLaMA 3.3 70B running on Groq, chained together using LangChain LCEL pipelines.

Beyond the app, langchain_groq.ipynb goes further building an autonomous agent that can search the web, look up Wikipedia, solve maths, and hold a conversation with memory.

---

🚀 The App — Restaurant Name Generator

Select a cuisine → get a restaurant name → get a full menu. Instantly.

Cuisine	Generated Name (example)	Menu Style
🇳🇬 Nigerian	Naija Royale	Jollof, Egusi, Suya Platter
🇮🇹 Italian	La Bella Cucina	Pasta, Risotto, Tiramisu
🇲🇽 Mexican	Casa del Sol	Tacos, Guac, Churros
🇸🇦 Arabic	Al Majlis	Shawarma, Hummus, Baklava
🇺🇸 American	The Liberty Grill	Burgers, Ribs, Cheesecake

How it works under the hood:

User picks cuisine
      ↓
Chain 1 → "Suggest a fancy restaurant name for {cuisine} food"
      ↓
Chain 2 → "Suggest menu items for {restaurant_name}"
      ↓
Streamlit displays results

Two prompts. Two chains. One clean output. That's LCEL sequential chaining.

---

🤖 The Agent — langchain_groq.ipynb

A standalone LangChain agent built from scratch that can:

• 🔍 Search the web via Google (Serper API)
• 📖 Look up Wikipedia for factual answers
• 🧮 Solve calculations with a built-in calculator tool
• 🧠 Remember conversations using buffer and window memory

The agent doesn't just answer, it thinks:

Question → Thought → Action → Observation → Final Answer

This is the ReAct reasoning pattern that powers modern AI assistants.

---

🛠️ Tech Stack

Tool	Purpose
🦜 LangChain	Chains, agents, memory, prompt templates
⚡ Groq	Blazing fast LLM inference
🤖 LLaMA 3.3 70B	The brain behind every response
🖥️ Streamlit	Web interface
🐍 Python 3.10	Core language

---

📁 Project Structure

langchain-groq-restaurant-agent/
│
├── 📂 RestaurantNameGenerator/
│   ├── main.py               # Streamlit UI
│   ├── langchain_helper.py   # LCEL chain logic
│   └── secret_key.py         # Your API keys go here
│
├── 📓 langchain_groq.ipynb   # Agent + memory implementation
├── 📄 requirement.txt
├── 📄 secret_key.py          # Root-level keys for the notebook
└── 📄 README.md

What was removed in this project because of streamlit hosting

1. In main.py

• Removed - import langchain.py

2. In langchain.py

• Removed - from langchain_groq import ChatGroq
• Converted CORRECT_PASSWORD = "your password" # Hardcoded in code to CORRECT_PASSWORD = st.secrets["APP_PASSWORD"] # Instead of hardcoded

---

⚙️ Setup & Run

1. Clone the repo

git clone https://github.com/Dominionai/langchain-groq-restaurant-agent.git
cd langchain-groq-restaurant-agent

2. Create a virtual environment

python -m venv ai_env

# Windows
ai_env\Scripts\activate

# Mac / Linux
source ai_env/bin/activate

3. Install dependencies

pip install -r requirement.txt

4. Add your API keys & password

Fill in secret_key.py:

groq_api_key = ""       # → console.groq.com (free)
serper_api_key = ""     # → serper.dev (free tier)
APP_PASSWORD = ""       # → add your password

5. Run the app

cd RestaurantNameGenerator
streamlit run main.py

---

💡 Core Concepts Implemented

Concept	Where
PromptTemplate + LCEL pipe | operator	langchain_helper.py
Sequential chaining (output → next input)	langchain_helper.py
StrOutputParser	langchain_helper.py
Tool-calling agent (Search, Wikipedia, Calculator)	langchain_groq.ipynb
ReAct reasoning loop	langchain_groq.ipynb
ConversationBufferMemory	langchain_groq.ipynb
ConversationBufferWindowMemory	langchain_groq.ipynb

---

🔮 Coming Next

• Deploy to Streamlit Cloud (live demo link)
• Add custom cuisine input by the user
• RAG project - chat with your own documents
• Multi-agent system with LangGraph

---

👤 Author

Egwuatu Chibuike Dominion AI Engineer | Nnamdi Azikiwe University

---

⭐ If this was useful or interesting, a star means a lot — thank you!