Hello! I've been working on building a flink alternative using DuckDB.<p>Building off of DuckDB makes it easy to leverage all the integrations DuckDB supports.<p>Using batching it's trivial to insert 5000 rows / second into a small postgres instance running locally in docker!<p>Would love your thoughts and feedback, thank you!<p>What do your data stacks look like for doing this non-differentiating data work like connecting streams and sources together?