80
It all began with a question:
"What if analyzing complex data & getting visual understanding was as easy as chatting with a friend?"
That question sparked the birth of Analyza — our Minor project 2 aimed at transforming how users interact with data.
🔧 With teammates Dhuruv Kumar and Khushi Chauhan, and mentorship from Dr. Deepak Sharma, we spent over four months designing a system that bridges the gap between raw data and real insights.
So, what is Analyza?
A smart data analysis and visualization platform where users can:
📂 Upload CSV files
💬 Ask questions in plain English
📊 Instantly receive insights and visualizations — customized to their needs
The core modules of Analyza include:
: Handles descriptive analytics using language models
: Offers predictive and diagnostic analytics built with FastAPI and Python libraries like Pandas, NumPy, Scikit-learn
: Users can generate and personalize visualizations — choose graph types, color schemes, and variables to make their insights more engaging
📈 Smart Suggestions = Smarter Results
One of our favorite features? Analyza intelligently suggests the best-suited features and targets based on your dataset — helping users get better accuracy and insights without being data science experts.
🔍 Whether you want to run a correlation matrix, detect anomalies, or perform root cause analysis, Analyza’s diagnostic module has your back.
📉 For prediction? Just choose your target and algorithm, and get clear, actionable outputs.
🧠 And for descriptive summaries? The LLM module breaks it down in simple language, along with visual insights.
👩💻 Tech Stack Breakdown:
Frontend: React.js
Backend: FastAPI
ML & Analytics: Pandas, NumPy, Scikit-learn, Seaborn, Matplotlib
LLM Integration: Gemini 2.0 Flash
🌐 Who’s it for?
Analyza is designed to support:
🎓 Educators and students analyzing academic performance
🏢 Businesses evaluating growth and strategy
🏥 Healthcare professionals predicting trends
📊 Researchers exploring complex datasets
We didn’t just build a tool. We built an experience — where users don’t just look at data but understand it.
Built with