[EN] Motherduck / DuckDB Special - multiple talks
Details
Event language: EN
🎤 Motherduck / DuckDB Special
⏳ Agenda:
* 1800: Doors open + Snacks 🍕 + Drinks 🍹 + Networking 💬
* 1900: Talk 1: Mehdi Ouazza (Motherduck) [about 30mins]
* 1945: Talk 2: Nicolas Renkamp (Merck) [about 30mins]
* Afterwards until around 2200: Q&A, Networking, Drinks
📝 Session abstracts
Talk 1 - Mehdi Ouazza (Motherduck): Vibecoding for Data Engineers: AI, DuckDB, and the Joy of Fast Feedback
Tired of copy-pasting broken SQL from ChatGPT into your data warehouse and manually debugging cryptic errors? There’s a better way. In this talk, we’ll explore a new, high-feedback coding workflow for data engineers that combines AI tooling, DuckDB, and MotherDuck to make building, testing, and refining queries feel fast, safe, and—dare we say it—fun.
Talk 2 - Nicolas Renkamp (Merck): Migrate from PySpark to DuckDB - without code changes
Engines like duckdb or libraries like SQLFrame offer a compatibility layer to run existing pyspark code without Spark and migrate the compute to alternative query engines that promise an X-fold increase in performance for small to medium data sets at lower costs. In this session, I give an overview of the current offerings in the open source space and share our successes and failures from migrating thousands of Spark builds to DuckDB.
