Comment by estetlinus

I love what I am hearing. I still see a lot of engineers using pandas, but it is such a horrible tools. You usually find an abandoned notebook with 100s of ”df_final_2” with sequential wrangling, making it impossible to understand what’s happening. Notebooks are also horrible for the same reason IMO.

With chat-your-data you have Hex, Claude + MCP, snowflake, Databricks etc… everyone’s in on it.