Last week Seattle, this week Prague, DuckDB user meetups are happening with many great talks! I'll be sharing them soon on YouTube.
Thanks for hosting this in Prague GoodData !

A common pattern for ensuring data quality is the 'Write-audit-publish' method. You essentially only serve the final dishes once a data taster has checked for any 'poison' and validated them 🧑‍🍳.

This one is going to be fun!!

Often, the data team is the last to be aware of an issue, usually hearing about it from consumers.
If you can't react before they find out, you lose a bit of trust. Probably forever.
Having good data observability is key to preventing this.

Maybe a new career path

Prague here I come!

Can LLMs be useful for data processing?
While we can store useful common prompts, how do we handle repeatability and costs at scale? 🤔

I just checked the first commit message of 'Git.' No kidding, it's giving me some nightmares.

Table war is still on, but it's a step forward for folks to easily switch between execution engine depending on the need.

It's not news that StackOverflow traffic has drastically decreased since ChatGPT's release.
And if you can't beat them, join them?
stackoverflow.co/company/press/…

TLDR ChatGPT's latest release: Itzme – speed and versatility unleashed.

I def want to get my hands on LLMs for web scraping, and I found ScrapeGraphAI. They have pretty decent documentation, let's see 👀github.com/VinciGit00/Scr…

SQL vs DataFrame.
Usually use DataFrame (w/ Python) for all ingestion work, and SQL for most of the transformation.

Join Brian Olsen from Tabular & Mehdi Ouazza for a fun, informative session diving into table formats, with a special focus on Iceberg & DuckDB.
Explore the latest advancements & enjoy a hands-on demo showcasing the power of these tools.
Get ready to quack… and code!

Scare a data engineer during an interview in one line 💬
'Our data sources are primarily CSVs and Excel sheets hosted on an FTP server.'

Tech vs. traditional companies: IT as a profit center vs. an expense.
Likewise, product teams vs. data teams: revenue drivers vs. cost centers.