Skip to content

README.md

Latest commit

 

History

History
60 lines (41 loc) · 1.88 KB

README.md

File metadata and controls

60 lines (41 loc) · 1.88 KB

rill-project-build-helper

A starter Rill project that includes a recommended folder structure, example files, and guidance for connecting data sources.

Overview

This repository provides a boilerplate Rill project you can use as a starting point for analytics development. It includes:

  • A pre-built folder structure (models/, sources/, dashboards/, connectors/, etc.)
  • Sample configuration files
  • Inline documentation to help you understand how projects are organized
  • Guidance for choosing your compute engine and connecting to data

Rill lets you build real-time dashboards powered either by its managed embedded OLAP engine (DuckDB or ClickHouse) or by connecting live to external OLAP systems like MotherDuck, ClickHouse, or Druid.

Repository Layout

This repo contains two example project setups, one for each compute mode:

rill-project-build-helper/
├── rill-managed/      # Use this for Option A (embedded DuckDB/ClickHouse)
├── live-connector/    # Use this for Option B (Live connect to external OLAP)
└── README.md

Getting Started

1. Choose your compute mode

Rill supports two execution models:

Option A: Managed Embedded Engine (default)

Rill runs a local managed DuckDB or ClickHouse instance.

Advantages:

  • Extremely fast in-memory queries
  • Real-time interactions without burdening your warehouse
  • Reduced cloud compute costs
  • Zero operational overhead

This is the mode used in this project.

Option B: Live Connect to an External OLAP Engine

Examples: MotherDuck, ClickHouse, Druid

In this mode, Rill does not ingest or store your data — instead it queries your OLAP engine directly.

Advantages:

  • Centralized governance
  • Leverage an already managed cluster
  • Query performance dependent on your OLAP engine

🙌 Contributing

Feel free to fork, open issues, or submit PRs to improve the project template.