Skip to content

Repository files navigation

Header

Typing SVG


Status Python Groq Streamlit Voice GenAI


Live App LinkedIn GitHub


🤖 About KAUTILYA

kautilya = {
    "name"       : "KAUTILYA — AI-Powered Data Analysis Agent",
    "named_after": "Chanakya (Kautilya) — Ancient Indian Strategist & Master of Intelligence",
    "ai_model"   : "Groq LLaMA3 70B — Fastest AI Inference Available",
    "voice"      : ["Hindi", "English"],
    "input"      : ["CSV", "Excel", "Text Questions", "Voice Questions"],
    "output"     : ["AI Insights", "Auto Visualizations", "Voice Responses"],
    "deployment" : "Live Streamlit App",
    "unique"     : "Agentic AI — thinks, analyzes & recommends autonomously"
}

🧠 "Named after the ancient Indian strategist Chanakya — KAUTILYA doesn't just answer questions. It thinks, analyzes, and delivers intelligence."


🎬 Live Demo

KAUTILYA Live Demo


🖥️ App Screenshots

Complete Interface Overview

KAUTILYA Collage

Main Interface — Home Screen

KAUTILYA UI

Dataset Loaded — 9,994 Rows

Dataset Loaded

AI Answering Questions

Question Asked

Auto Visualization — Correlation Heatmap

Visualization

Voice Input & Speak Button

Speak Button


✨ Key Features

Feature Details
🧠 AI Brain Groq LLaMA3 70B — transformer-based LLM
🎤 Voice Input Hindi & English via Google Speech API
🔊 Voice Output Google Text-to-Speech (gTTS)
📊 Auto Analysis Complete dataset analysis in one click
📈 Visualizations Distribution, Correlation, Categories, Missing Values
🔍 Data Explorer Preview, Statistics, Column Info
🔐 Security API key stored in .env — never exposed
📁 Sample Datasets E-Commerce Orders, Superstore Sales

🤖 What Makes This Agentic AI?

Traditional Chatbot    = Answers predefined questions
KAUTILYA (Agentic AI)  = Autonomously:
                         → Loads and understands any dataset
                         → Maintains conversation memory
                         → Generates insights without instructions
                         → Recommends next steps proactively
                         → Responds in voice — Hindi & English

📊 Sample Datasets Included

Dataset Rows Domain
Order Details 500 E-Commerce
Sample Superstore 9,994 Retail Sales
NIFTY50 (Download separately) 2,35,192 Stock Market

⚠️ NIFTY50 dataset not included due to GitHub 25MB limit. Download from Kaggle and place in root folder.


🚀 Run Locally

git clone https://github.com/DeveshShukla23/KAUTILYA-AI-Data-Analysis-Agent.git
cd KAUTILYA-AI-Data-Analysis-Agent
pip install -r requirements.txt
echo "GROQ_API_KEY=your_groq_api_key_here" > .env
streamlit run agent.py

Get your free API key at: console.groq.com


💡 How to Use

1. Run the app locally or open Live App
2. Select data source from sidebar:
   → Upload your own CSV/Excel file
   → Or use provided sample datasets
3. Ask questions in text or voice
4. Click "Auto Analyze Dataset"
   for complete instant analysis
5. Explore visualizations in
   "Auto Visualizations" tab

🔐 Security

API Key stored in .env file
.env listed in .gitignore
= API key NEVER pushed to GitHub
= Safe for public repositories

📂 Project Structure

KAUTILYA-AI-Data-Analysis-Agent/
│
├── agent.py                          → Main Streamlit application
├── requirements.txt                  → Python dependencies
├── .gitignore                        → Excludes .env file
├── kautilya_demo.gif                 → Live demo animation
├── KAUTILYA_LinkedIn_Collage.png     → App overview collage
├── Kautilya_UI2.png                  → Main interface screenshot
├── Kautilya Load Dataset.png         → Dataset loaded screenshot
├── kautilya Question asked.png       → AI answering screenshot
├── Kautilya visualization.png        → Auto visualization screenshot
├── kautilya speak button.png         → Voice controls screenshot
│
└── Sample Data/
    ├── Order Details.csv             → E-Commerce dataset
    └── Sample - Superstore.csv       → Retail sales dataset

🛠️ Tech Stack

Python Groq Streamlit SpeechRecognition gTTS Pandas Matplotlib Plotly


🔮 Future Improvements

🔄  Real-time database connectivity
🤖  Automated ML model training on uploaded data
📄  PDF report generation
🧠  Multi-step agent planning using LangChain
🌐  Full Hindi/Hinglish language support
📊  Dashboard mode for business reporting

💡 Skills Demonstrated

Agentic AI Voice Tech Data Analysis
✅ LLM Integration ✅ Speech Recognition ✅ Auto EDA
✅ Groq API ✅ Text-to-Speech ✅ Auto Visualizations
✅ Conversation Memory ✅ Hindi & English ✅ Dataset Intelligence
✅ Autonomous Insights ✅ Voice Controls ✅ Business Recommendations

👨‍💻 Author

Devesh Shukla Data Analyst | Data Scientist | Agentic AI Builder

LinkedIn GitHub Live App CHANAKYA


If you find this useful, please give it a star!

Footer

About

🤖 Agentic AI Data Analysis Agent | Groq LLaMA3 70B | Voice Enabled Hindi & English | Auto Visualizations | Live Streamlit App

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages