Docs · Get an API key · Quick Start
Welcome to our end-to-end examples! Each project is self-contained and walks you through a different capability - from building your first pipeline to production patterns like write-audit-publish and LLM integration.
Every example assumes the following setup:
- Bauplan account - sign up at bauplanlabs.com and get your API key.
- CLI and SDK - install the Bauplan CLI and Python SDK. See the installation docs.
- uv - we use uv for Python dependency management. Inside any example folder, run
uv syncto install its dependencies. - Familiarity with the basics - if you're new to Bauplan, start with the quick start tutorial.
Bauplan operates entirely in the cloud. Your local scripts communicate with the platform, which orchestrates and executes data workflows on serverless infrastructure - there is nothing to provision or manage.
Some examples require additional services (Prefect, Pinecone, MongoDB, an LLM API key, AWS credentials). These are noted in the individual READMEs.
These examples introduce core Bauplan concepts one at a time - pipelines, data quality, branching, safe ingestion, and more. Each one focuses on a specific capability so you can learn the platform incrementally.
| # | Example | What you'll learn |
|---|---|---|
| 01 | Pipeline to Dashboard | Build a two-model pipeline over NYC taxi data and visualize the output in a Streamlit dashboard |
| 02 | Data Quality & Expectations | Add expectation tests to catch data quality issues before they reach production |
| 03 | Safe Ingestion on a Schedule | Implement the Write-Audit-Publish (WAP) pattern with Prefect to safely ingest data on a schedule |
| 04 | Using LLMs with Secrets | Ingest PDFs from S3, extract structured financial data with an LLM, and explore results in Streamlit |
| 05 | Advanced Git for Data | Deep dive into branching, time travel, tagging, reverts, fault isolation, and multi-step transactions |
| 06 | Data Engineering with an AI Coding Assistant | Interactive narrative: build a production telemetry pipeline with an AI coding assistant and Bauplan skills |
Full applications that combine Bauplan with third-party tools and services. Find them in the projects/ folder.
| Example | Description |
|---|---|
| RAG Service Support Agent | RAG pipeline over Stack Overflow data with Pinecone vector search and LLM-powered Q&A |
| Playlist Recommender | Embedding-based music recommendations with MongoDB Atlas vector search |
| From Notebooks to Prod | From a marimo notebook to a production pipeline - same Python functions, no rewrite |
| Medallion for Telemetry Data | Bronze-Silver-Gold medallion architecture for sensor telemetry with DuckDB, Polars, and a dashboard |
- Bauplan documentation
- Architecture paper — the design behind Bauplan
- Ergonomics paper — developer experience and API design
- bauplanlabs.com — product overview
All code in this repository is released under the MIT License. Third-party tools and services used in the examples (Prefect, Pinecone, MongoDB, etc.) are subject to their own licenses.