👀Sounds interesting...

Useless, and at the same time quite judgemental.

There are references to technical concepts like APIs, data, and user profiles, but it is unclear how these relate to the overall discussion. The participants use a lot of jargon and abbreviations, making it difficult to follow the flow of the conversation."

Has Java something akin to Scala's copy() in case classes? In Scala I'd set the fields in question to the same defined value.

I built a couple of smaller tasks using Junie, JetBrains' coding agent. For one, it takes too much time to produce code. Then you still have to either improve the code or ask it again to do it for you. And finally, it eats too much your AI quota to be used every day.

*it not I - gief edit!

And not only could it not figure it out, it sent me on a completely wrong and distracting detour.

I'm not convinced by 'vibe coding' at all.

Now we have to spend time to figure out if information about technology is real or AI slop 🙈

I thought the idea was to write tooling in Rust for performance and use Python for the API/SDK? But agree, efficiency is the least of Data Engineering’s problems. A growing zoo of tools is.

Nice!

Interesting, I'm also looking for alternatives to Databricks and their dependency-hellish runtimes. But being tied to one specific cloud vendor is also meh.

I remember using Keras (1.x I guess) from Scala. Is it Python-only now?

🎉