Cube.js

Cube.js

An open source analytics framework 📊

93 followers

Visit website
•
•
•
Cube.js is an open-source, modular framework for building analytics into your web apps. You can use Cube.js for internal business intelligence tools, or customer-facing analytics.
This is the 2nd launch from Cube.js. View more

Cube.js Web Analytics

Build your own Google Analytics alternative
Cube.js Web Analytics was ranked #5 of the day for March 31st, 2020
Use this template to build your own open-source Google Analytics alternative. It's developer friendly, hackable, and embeddable. It is easy to install the full working application and then customize every part of it from data collection to and visualizations.
Cube.js Web Analytics gallery image
Cube.js Web Analytics gallery image
Cube.js Web Analytics gallery image
Cube.js Web Analytics gallery image
Launch tags:
Open Source•Analytics•Developer Tools
Launch Team

What do you think? …

Artyom Keydunov
Maker
Artyom and Pavel from Cube.js here - We've been receiving a lot of questions and requests from the community on how to build a web analytics platform, similar to Google Analytics, with Cube.js. That led us to building a ready-to-use application and template for this, which everyone can easily install and customize. By building and hosting such an application yourself, you'll be able to fully control the privacy and don't need to share your users’ data with 3rd parties. It also means no vendor lock-in and you completely own the underlying data. We've designed this template to be fully hackable and embeddable. This means it's very easy to customize on every level—from data collection to metrics definitions and visualizations. The frontend is a pure React application based on Material UI without any custom styles. You can embed any part of the frontend into your existing application and customize the look and feel to match your styles. Last, but not least, we've focused a lot on the performance to keep response time under 50 ms and make it scale up to several million daily active users. We hope you enjoy it and build a lot of great and useful applications with this template. Happy hacking!
Dmitry Petrov
@keydunov That sounds really cool. It should be much more flexible then g-cloud and mixpanel. A couple of questions: 1. does it support regular stuff like cohort analysis? 2. I run my own DB, right? Any support for custom DB queries? 3. If custom queries are supported, can I go a bit further and score data against my ML models? How am I supposed to serve the model?
Artyom Keydunov
Maker
@fullstackml Thank you! Since it's not a monolithic application, but rather a template designed for customization on any level it is easy to add all these custom things. 1. The template itself doesn't have it, but it is very easy to add with Cube.js data schema, which powers metrics definitions. 2. You can customize all the queries and add new ones but changing Cube.js data schema definitions. 3. We don't go into ML space directly, but I believe it should be possible to create data schema against the tables in the database populated with the results of ML models.
Sm0ke
Good luck. Cube.js is a great tool.
Artyom Keydunov
Maker
@achirilov Thank you!
Sm0ke
@keydunov Yw. Cube.js fill a huge gap in the visual analytics field. Also, your (Slack) community traction is impressive. Keep up the good work.
Nikita Korotaev
Hi @keydunov! Congrats on the great demo! Curious what are the main differences with Matomo?
Artyom Keydunov
Maker
@nikitakorotaev Thank you! It is a great question. Unlike Matomo, it is not a monolithic application, but rather a set of modules, which based on different established open-source technologies: Snowplow, PrestoDB, Cube.js, and React with Material UI on frontend. You can change or replace any component with technology that better fits your needs. Additionally, due to Cube.js data schema you can customize how metrics and dimensions are defined in a case the default definitions don't work for you.