InsightFlow
InsightFlow is an interactive data analytics platform built with Python, Streamlit, FastAPI, and PostgreSQL.
The application allows users to upload CSV datasets and automatically generate:
- business insights,
- KPIs,
- statistical summaries,
- interactive visualizations,
- exploratory data analysis.
The project uses a modular frontend/backend architecture with REST API communication and persistent dataset storage using PostgreSQL.
⸻
Live Demo
https://insightflow-lucia-hernandez.streamlit.app/
Note: the backend runs on Render's free tier, so the first request after a period of inactivity can take up to a minute to respond while the server wakes up.
⸻
GitHub Repository
https://github.com/luciaaherr/InsightFlow
⸻
Features
- CSV upload and processing
- Automatic dataset analysis
- Dynamic KPI generation
- Interactive Plotly visualizations
- Missing values detection
- Statistical summaries
- Automatic numerical and categorical column detection
- FastAPI backend architecture
- PostgreSQL database integration
- Persistent dataset history
- REST API endpoints
- Swagger API documentation
- Modular project structure
- Frontend/backend separation
⸻
Technologies Used
Frontend
- Streamlit
- Plotly
Backend
- FastAPI
- SQLAlchemy
- PostgreSQL
- Python
- Pandas
Development Tools
- Git
- GitHub
- Python Virtual Environment (venv)
Deployment
- Streamlit Community Cloud
⸻
Project Architecture
Frontend (Streamlit) ↓ REST API Requests ↓ Backend (FastAPI) ↓ SQLAlchemy ORM ↓ PostgreSQL Database
⸻
Project Structure
InsightFlow/ ├── frontend/ │ ├── app.py │ ├── visualizations.py │ └── requirements.txt │ ├── backend/ │ ├── main.py │ ├── database.py │ ├── requirements.txt │ │ │ ├── routes/ │ │ ├── datasets.py │ │ ├── kpis.py │ │ ├── statistics.py │ │ └── business_insights.py │ │ │ ├── models/ │ │ └── dataset.py │ │ │ └── services/ │ ├── analysis_service.py │ ├── kpi_service.py │ ├── statistics_service.py │ ├── business_insights_service.py │ ├── business_metrics.py │ └── csv_loader.py │ ├── requirements.txt ├── README.md ├── .gitignore └── .env ⸻
Current API Endpoints
Datasets
- POST /datasets/insights
- GET /datasets/history
KPIs
- POST /datasets/kpis
Statistics
- POST /datasets/statistics
Business Insights
- POST /datasets/business-insights
⸻
How to Run Locally
Clone repository
git clone https://github.com/luciaaherr/InsightFlow.git
Enter project folder
cd InsightFlow
Create virtual environment
python3 -m venv venv
Activate virtual environment
Mac/Linux
source venv/bin/activate
Windows
venv\Scripts\activate
Install dependencies
pip install -r requirements.txt
⸻
PostgreSQL Setup
Install PostgreSQL locally and create the database:
CREATE DATABASE insightflow_db;
The backend uses SQLAlchemy ORM to connect to PostgreSQL and store uploaded dataset metadata.
⸻
Run Backend
./venv/bin/python -m uvicorn backend.main:app --reload
Backend documentation:
⸻
Run Frontend
./venv/bin/python -m streamlit run frontend/app.py
⸻
Future Improvements
- User authentication with JWT
- Multi-user dataset management
- Cloud PostgreSQL deployment
- Docker support
- Advanced analytics engine
- Export reports to PDF
- Dataset dashboards
- AI-generated insights
- Background processing system
⸻
Author
Lucía Hernández Computer Engineering Student — Universidad de la República (Udelar)