DuckCon #3
who it's for
DuckDB users and contributors, data engineers, analytics engineers, Python data scientists, dbt practitioners, log analytics and observability engineers, developer-tooling and IDE developers
topics
DuckDB attach SQLite backend; parallel JSON autodetect; parquet partitioned writes; pivot and lateral joins; out-of-core hash aggregate spilling; Apache Iceberg read support; Malloy single-query compilation; nested JSON implicit joins; dbt adapter for DuckDB; column-level lineage from AST; pandas-to-SQL pushdown (Ponder); Lance columnar multimodal format
| title | author | length |
|---|---|---|
| Intro | Hannes Mühleisen | 03m |
First US event lands in San Francisco for DuckCon with a status update on building a database framed as “a very stupid idea” that will take ten years, a… | ||
| State of the Duck | Mark Raasveldt | 27m |
Releases 0.7 and 0.8 deliver concrete features: attach + SQLite backend lets DuckDB attach and transact against SQLite files; parallel JSON autodetect reads… | ||
| ‘Data is Rectangular’ and Other Common Misconceptions | Lloyd Tabb | 23m |
Claims that data is a graph not rectangular and that Malloy stores join topology so aggregates always correct regardless of join pattern. Malloy compiles each… | ||
| DuckDBT: Not a database or a dbt adapter but a secret third thing | Josh Wills | 22m |
Adapter glues DBT to DuckDB so DBT runs against local DuckDB instances and object stores without standing up a cloud warehouse; it supports Iceberg, SQLAlchemy… | ||
| Watershed: Get to zero carbon, fast | Jessica Zhu | 05m |
Watershed runs monthly carbon-footprint parquet files (largest customer 17M rows, typical footprint ~750 MB) from Google Cloud Storage through hosted DuckDB… | ||
| Building a SQL editor around “fast” | Hamilton Ulmer | 06m |
Rill Data pivots from Druid to DuckDB and builds a SQL IDE that leverages sub-second query latency to remove the no run button — results start appearing as… | ||
| KalDB: Log analytics with DuckDB | Suman Karumuri | 07m |
KalDB adds a plugable storage engine so one logging system serves both search and analytics by combining Lucene for search with DuckDB for fast analytics; it… | ||
| Pandas on DuckDB with Ponder | Aditya Parameswaran | 05m |
Ponder translates pandas to SQL and pushes computations down to your warehouse, producing large generated queries (example: 300+ lines for a min‑max… | ||
| Bringing AI to DuckDB with Lance columnar format for multi-modal AI | Chang She | 05m |
Lance columnar format and LanceDB aim to replace Parquet+TFRecords copies by storing embeddings, images, text, point clouds and nested annotations together… | ||
| nothing matches | ||