How neha-lodh/insightflow works
InsightFlow is an AI-assisted data analysis app with two user-facing modes: a Flask-served browser interface and a Streamlit interface. The main workflow is to sign in, upload a dataset, inspect its profile and quality, ask questions or explore charts, then download a report. The Python engine supports cleaning, profiling, query planning, safe analysis and report generation; account handling is a separate subsystem. The browser UI implementation was not sampled, so its documented capabilities are shown without inferring its internal calls. Connections are limited to excerpted code or README statements.
Main entry points: app.py Streamlit App · insightflow_ui.html Web Dashboard · server.py Flask Server · auth.py Authentication
Components 15
User Interfaces
- Streamlit App
app.py - Web Dashboard
insightflow_ui.html - Explore And Quality
insightflow_ui.html - Questions And Charts
insightflow_ui.html - Reports
insightflow_ui.html
Data Analysis
- Analysis Engine
core.py - Question Planning
core.py - Safe Analysis
core.py
Accounts
- Sign-In Page
login.html - Authentication
auth.py - Account Database
users.py
Web Service
- Flask Server
server.py - Excel Import
server.py
Other components
- Analyst
- Browser
Connections 21
- Analyst → Streamlit App: uses
README.md - Analyst → Browser: uses
README.md - Browser → Web Dashboard: opens
README.md - Flask Server → Web Dashboard: serves
server.py - Browser → Excel Import: uploads Excel
README.md - Excel Import → Flask Server: handled by
server.py - Flask Server → Authentication: registers routes
server.py - Sign-In Page → Authentication: uses auth endpoints
- Authentication → Account Database: validates and authenticates
auth.py - Streamlit App → Account Database: registers and signs in
app.py - Streamlit App → Analysis Engine: prepares dataset
app.py - Streamlit App → Question Planning: plans questions
app.py - Streamlit App → Safe Analysis: executes analysis
app.py - Streamlit App → Analysis Engine: builds reports
app.py - Web Dashboard → Explore And Quality: provides
README.md - Web Dashboard → Questions And Charts: provides
README.md - Web Dashboard → Reports: provides
README.md - Analyst → Explore And Quality: inspects data
README.md - Analyst → Questions And Charts: asks questions
README.md - Analyst → Reports: downloads
README.md - Question Planning → Safe Analysis: creates plan
core.py