Python and SQL: Better Together, Powered by DuckDB
summary
DuckDB claims a fast, in‑process analytical engine that integrates with Python dataframes and common databases. It runs in-process (no socket roundtrips), is MIT licensed and a single pip install with no external deps, and can read Pandas Polars Apache Arrow in place. The talk shows it handles larger‑than‑RAM workflows, offers relational/Ibis/PySpark APIs, reads/writes Postgres MySQL SQLite and cloud object storage, and reports 5 queries on 1B rows in ~24s.
AI-written summary · about
chapters (10)
- 0:00Introduction to DuckDB
- 2:38Bridging SQL and data science
- 4:58What is DuckDB
- 9:05Friendlier SQL features
- 11:47Integration with dataframes
- 16:32Pythonic APIs
- 19:55Using Ibis and PySpark
- 22:06Performance benchmarks
- 25:22Ecosystem and versatility
- 28:20Q&A session
related talks
30mDuckDB: Bringing analytical SQL directly to your Python shell
27mIn Process Analytical Data Management with DuckDB
34mData Wrangling with DuckDB
25mProcessing Biggish Data with DuckDB and Python
01h 02mAnalytics for not-so-big data with DuckDB
28mAnalyze data easily with duckdb - and the implications on data architectures
40mDuckDB: Harnessing in-process analytics for data science and beyond
05mDuckDB Powering Interactive Notebooks