GitHub user seongjinyoon added a comment to the discussion: Ambient "operator recommender" — predictive next-operator suggestions on the canvas
Hi @gupta-sahil01 @chenlica I worked on Texera operator embeddings. The motivating problem was that most approaches ignore PythonUDF operators (since they are user-defined and don't have fixed semantics) so I trained an embedding model that incorporates them when ranking next-operator suggestions. On evaluation: it reached ~75% top-3 hit rate (image below). <img width="493" height="349" alt="Screenshot 2026-05-28 at 10 29 35 AM" src="https://github.com/user-attachments/assets/469524bd-834a-40a5-99f5-1c3027e00f21"; /> GitHub link: https://github.com/apache/texera/discussions/5240#discussioncomment-17094886 ---- This is an automatically sent email for [email protected]. To unsubscribe, please send an email to: [email protected]
