title	emoji	colorFrom	colorTo	sdk	sdk_version	app_file	pinned
Smart GPT with Tools	🤖	blue	green	streamlit	1.32.0	frontend.py	false

🤖 Smart GPT with Tools – Intelligent LangGraph Chatbot

An intelligent chatbot built with LangGraph that uses conditional tool routing, reflection-based validation, and multi-tool reasoning for high-quality responses.

✨ Features • 🏗️ Architecture • 📁 Project Structure • 🚀 Quick Start • 🔬 How It Works

---

�🎯 Overview

Smart GPT with Tools is a LangGraph-powered conversational AI system that intelligently decides:

• when to answer directly,
• when to use external tools,
• and when to reflect and improve its own responses.

Unlike linear chat pipelines, this system uses graph-based control flow to route queries through tools, validation, and refinement stages before delivering the final answer.

The result is a chatbot that is faster, more accurate, and more reliable.

---

💡 Why This Project

Traditional chatbots:

• Call tools unnecessarily
• Return unverified answers
• Lack self-correction

This project demonstrates how to build:

• conditional reasoning flows,
• reflection-based quality assurance,
• and tool-augmented intelligence

It closely resembles production-grade agent systems used in real-world GenAI applications.

---

✨ Features

🎯 Intelligent Routing

• Conditional tool execution using LangGraph
• Direct LLM responses for simple questions
• Automatic decision-making by the model

🔧 Multi-Tool Capabilities

• DuckDuckGo web search for real-time information
• Calculator for mathematical expressions
• Easily extensible tool interface

🔍 Quality Assurance

• Reflection node validates every response
• Improves clarity, correctness, and completeness
• Prevents low-quality or partial answers

💬 Advanced Chat Experience

• Conversation history with memory
• Multi-threaded chat sessions
• Persistent state using checkpointer
• Streaming token-by-token responses

---

🏗️ Architecture

High-Level Flow

User Query
│
▼
Chat Node
│
▼
Conditional Router
│
├──► Tool Execution (Search / Calculator)
│           │
│           ▼
│       Chat Node
│
└──► Reflection Node
│
▼
Final Response

Why LangGraph

LangGraph enables:

• conditional edges
• tool-aware routing
• validation loops
• clean separation of responsibilities

---

🔬 How It Works

1. Chat Node

Receives user input and invokes the LLM with tool bindings.

2. Conditional Router

Inspects the LLM output:

• tool call detected → route to tool execution
• no tool needed → route directly to reflection

3. Tool Execution

Executes requested tools such as:

• web search
• calculator

4. Reflection Node

Evaluates the response for:

• correctness
• clarity
• completeness

Improves the answer if needed before returning it.

5. Memory Management

• Thread-based conversation separation
• Full chat history retained per session
• Supports multi-user and multi-chat workflows

---

🛠️ Tech Stack

Component	Technology
Orchestration	LangGraph
LLM Framework	LangChain
LLM Provider	Groq
Web Search	DuckDuckGo
UI	Streamlit
Language	Python

---

🚀 Quick Start

Prerequisites

• Python 3.8+
• Groq API key

Environment Setup

Create a .env file:

GROQ_API_KEY=your_groq_api_key

Install Dependencies

pip install -r requirements.txt

Run the App

streamlit run frontend.py

The app will open at http://localhost:8501.

---

📁 Project Structure

Smart_Gpt_with_tools/
├── src/
│   ├── states/
│   │   └── state.py              # ChatState definition
│   ├── llms/
│   │   └── llm.py                # LLM + tool setup
│   ├── nodes/
│   │   └── node.py               # Chat, router, reflection nodes
│   └── graphs/
│       └── graph_builder.py      # LangGraph construction
├── backend.py                    # Chatbot initialization
├── frontend.py                   # Streamlit UI
├── requirements.txt              # Dependencies
├── pyproject.toml                # Project config
└── README.md

---

🧪 Example Queries

Direct response:

• What is Python?
• Explain machine learning

Tool-based:

• What is 456 * 789?
• Latest news about AI
• Current weather in New York

---

🔧 Customization

Change LLM Model

Edit src/llms/llm.py:

llm = ChatGroq(model="your-model-name")

Add New Tools

@tool
def custom_tool(input: str) -> str:
    return result

Modify Routing Logic

Edit route_tools() in src/nodes/node.py.

Improve Reflection Logic

Customize prompts in reflect_node().

---

🚀 Deploy to Hugging Face Spaces

1. Create a new Space → SDK: Streamlit
2. Upload this repository
3. Add GROQ_API_KEY in Space Secrets
4. App runs automatically using frontend.py

---

📊 Why This Project Stands Out

• Not a basic chatbot
• Demonstrates agentic reasoning
• Uses reflection and validation
• Clean, scalable architecture
• Directly aligned with real-world GenAI systems

---

🔮 Future Improvements

• Tool result caching
• More specialized tools
• Confidence-based refusal
• Evaluation metrics dashboard
• Multi-agent extensions

---

👨‍💻 Author

Vivek Kumar Gupta AI Engineering Student | GenAI & Agentic Systems Builder

GitHub: https://github.com/vivek34561 LinkedIn: https://linkedin.com/in/vivek-gupta-0400452b6 Portfolio: https://resume-sepia-seven.vercel.app/

---

📄 License

MIT License © 2025 Vivek Kumar Gupta