A running list of things I’ve built or am building, mostly small, mostly in public.

---

waddler

A Go binary that reads a YAML pipeline file and runs a complete ETL pipeline using DuckDB as the query engine. No Python environment, no Spark cluster, no Airflow. Describe your sources, write your SQL, declare your output. Run waddler run pipeline.yml.

Supports CSV, JSON, Parquet, Postgres, and MotherDuck as sources. Writes Parquet, CSV, or directly to MotherDuck. Built-in validation rules and cron scheduling.

Built for organizations that have real data needs but not the resources for enterprise tooling: food banks, non-profits, small teams.

github.com/mehrabr/waddler  ·  Go, DuckDB, MotherDuck

---

spring-boot-starter-duckdb

In progress.

A Spring Boot autoconfiguration starter for DuckDB and MotherDuck, published to Maven Central. Zero-config DataSource auto-wired from application.yml, correct handling of DuckDB’s duplicate() connection model for read-write pools, Spring Boot Actuator health indicator, and MotherDuck token resolution.

The gap this fills: the official org.duckdb:duckdb_jdbc driver exists on Maven Central, but there is no Spring Boot starter for it. Every Java developer has to wire this manually and figure out DuckDB’s non-standard connection pooling rules themselves.

Go  ·  Java, Spring Boot, DuckDB, Maven Central