Insights is a tool to visually explore a PostgreSQL database, with an emphasis on generating graphs that show business performance over time.
Think of Google Data Studio or Google Looker, but totally free, self-hosted and without the "Google" part.
See a live demo for Widgets Inc, a fictional e-commerce site.
Please be aware that is an extremely early BETA release of Insights, which has not gone through any kind of security audit.
Use on a live server at your own risk!
To install, make sure you have Node 10+ installed and then run:
npm install -g insights
insights init
insights start
This creates a folder .insights
which contains all the config and runtime data.
- Self Hosted, installed via NPM
- PostgreSQL connection support
- Auto-detect your database schema, including all foreign keys!
- Connect to multiple databases
- Edit the schema and add custom SQL fields right there in the interface!
- Create subsets of your data (e.g. share only a few fields with marketing)
- Data explorer
- Filters on the data
- Time-based graphs
- Split the graph by some column (e.g. new users by country name)
- Keyboard navigation in the sidebar
- Saved views
- Pinned fields
- Embed React or <iframe> components and get data through the insights API
- Decent mobile support
- Log in with your Google Account
- Manage users in the interface
- Access control for subsets
- PDF and XLSX exports
- Way more and better graphs
- View generated SQL
- Dashboards
- Multiple lines from different sources on one chart in the dashboard
- Plugins?
Insights is MIT-licensed Sponsorware. If you use it in your business, please contribute towards its development!
To stay in touch and receive news when we release a significant update, please sign up here.
You can also follow me on Twitter to receive the latest updates.
If you want to help with development, run these steps:
# 1. fork the repo in github
# 2. clone it
git clone git@github.com:<YOUR_NAME_HERE>/insights.git
# 3. install all dependencies
cd insights
yarn
# 4. start the app
yarn run init
yarn start
# 5. open http://localhost:3000/ and hack away