Pilotis
Docs
Get Started
Foundations For a Smart Project
Open source ML projects generator, toolbox and conventions, designed for built-in quality
Do Less
Boilerplate
Generate your ML projects with a single command line, in minutes
Setup
Consistent toolbox and out of the box configuration designed to last for the entire lifetime of your project (jupyter plugins, build and delivery tools)
Team Noise
Opinionated lightweight conventions and tools over unhelpful polemics
Get More
Time
Immediately start to work on business concerns
Quality
Tests, dependency management, CD pipeline, data organization, best practices examples...
Hybrid Capabilities
Decoupled architectures and various modules for storage and computing, either locally, on the cloud, or both
Quick Start
1.  Install Pilotis
pip install pilotis
Copied to clipboard Copy
2.  Create your first project
pilotis init
Copied to clipboard Copy
What Pilotis is all about
Your ML project from scratch, production ready, in minutes
Generation of a clean, extensible, project skeleton
Your project is already hosted on Gitlab, with working CI/CD pipelines
Integrated Testing tools
Strongly opinionated conventions improving your teamwork
Data organization, for a clear process of data lifecycle
Fully managed virtual environments and dependencies
Collaborative Notebooks and easy integration with personal workspaces
Out of the box project architecture with generated code, tests, notebooks, and package management
Intuitive and extensible Apis
Helps you writing clean code
Facilitates an Hexagonal approach and a fully decoupled design
Guarantees compatibility with local and cloud datasets
Pre-installed and pre-configured set of tools
Notebooks
Libraries
Build tools
CLI