What is Dango?¶
Dango is an open-source data platform that integrates production-grade tools (dlt, dbt, DuckDB, Metabase) into a single, cohesive platform.
Works on your laptop today. Designed to scale to production tomorrow.
The Problem¶
Building a data platform typically requires:
- Weeks of setup and configuration
- Deep knowledge of multiple tools
- Complex infrastructure decisions
- Choosing between simple (limited) or complex (powerful) tools
Result: Data teams spend more time on infrastructure than analysis.
The Solution¶
Dango gives you a complete data stack with one command:
You get:
- dlt for data ingestion (29+ verified sources)
- dbt for SQL transformations
- DuckDB as your analytics database
- Metabase for dashboards and SQL queries
- Web UI for monitoring and management
Architecture¶
Dango uses a layered data architecture:
graph LR
A[Data Sources] --> B[dlt]
B --> C[Raw Layer]
C --> D[dbt]
D --> E[Staging]
E --> F[Intermediate]
F --> G[Marts]
G --> H[Metabase] Data Layers¶
- Raw - Immutable source of truth with metadata
- Staging - Clean, deduplicated data
- Intermediate - Reusable business logic
- Marts - Final business metrics
Tech Stack¶
| Component | Purpose | Why This Tool? |
|---|---|---|
| DuckDB | Analytics database | Embedded, fast, no server needed |
| dlt | Data ingestion | 29+ sources, schema evolution |
| dbt | Transformations | SQL-based, version controlled |
| Metabase | BI dashboards | Auto-configured, easy to use |
| Docker | Service orchestration | Consistent environments |
| FastAPI | Web UI backend | Fast, modern Python |
Core Features¶
Data Ingestion¶
- 29+ verified dlt sources (Stripe, Google Sheets, GA4, Facebook Ads, etc.)
- CSV upload and auto-sync
- Custom source development
- OAuth authentication for cloud sources
Transformations¶
- dbt auto-generation for staging models
- Full dbt project access
- SQL-based transformations
- Incremental model support
Monitoring¶
- Web UI with live pipeline status
- File watcher with auto-triggers
- Token expiry warnings
- Validation and health checks
Dashboards¶
- Metabase auto-configured with DuckDB
- Pre-built dashboard templates
- SQL query interface
- Dashboard backup and restore
Current Status: v0.0.5 (MVP)¶
What Works Now¶
- ✅ Full CLI with 10+ commands
- ✅ CSV, Stripe, Google Sheets, GA4, Facebook Ads sources
- ✅ dbt auto-generation for staging models
- ✅ Web UI with live monitoring
- ✅ Metabase dashboards
- ✅ File watcher with auto-triggers
- ✅ Custom sources via
dlt_nativetype
Coming Soon¶
- 🚧 v0.1.0: Google Ads, demo project, full documentation
- 🔮 Beyond v0.1.0: Cloud deployment guides, advanced scheduling, team collaboration
Design Philosophy¶
Dango is built chronologically - starting with local support for MVP, designed to scale to cloud production later.
Why This Approach?¶
- Get started fast - No cloud infrastructure needed
- Learn the tools - Master dlt, dbt, and SQL locally
- Scale when ready - Same tools, bigger infrastructure
- Production-grade - Best practices from day one
Target Users¶
- Solo data professionals - Complete stack, zero complexity
- Fractional consultants - Fast client onboarding
- SMEs - Analytics without engineering teams
- Learners - Production tools without production costs
Why Dango vs. Alternatives?¶
vs. Cloud Platforms (Snowflake, BigQuery)¶
| Dango | Cloud Platforms |
|---|---|
| ✅ Free (open source) | ❌ Pay per query |
| ✅ Runs on laptop | ❌ Requires cloud account |
| ✅ Fast iteration | ⚠️ Can be slow/expensive during dev |
| ⚠️ Local compute limits | ✅ Scales to petabytes |
vs. No-Code Tools (Fivetran, Airbyte Cloud)¶
| Dango | No-Code Tools |
|---|---|
| ✅ Fully customizable | ❌ Limited customization |
| ✅ Version controlled | ⚠️ UI-based configuration |
| ✅ Open source | ❌ Proprietary/paid |
| ⚠️ Requires code for complex cases | ✅ Point-and-click simple cases |
vs. DIY Stack¶
| Dango | DIY Stack |
|---|---|
| ✅ Integrated from day one | ❌ Weeks of integration work |
| ✅ Best practices built-in | ⚠️ Easy to make mistakes |
| ✅ Maintained by community | ❌ You maintain everything |
| ⚠️ Opinionated structure | ✅ Complete flexibility |
What Dango is NOT¶
- Not a SaaS platform - It's a CLI tool that runs locally
- Not cloud-only - MVP is local, cloud support coming later
- Not a BI tool - It integrates BI (Metabase) but focuses on data infrastructure
- Not production-ready for enterprise - Currently MVP (v0.0.5)
Next Steps¶
Ready to try Dango?
- Install Dango - Get set up in minutes
- Quick Start - Run your first pipeline
- Core Concepts - Deep dive into architecture
Questions?¶
- GitHub: github.com/getdango/dango
- Issues: github.com/getdango/dango/issues
- PyPI: pypi.org/project/getdango