Mitch H.

Ali - Sorry for this, it looks like we've got a bug on our hands. I've escalated this to our engineering team to help look into it. We'll follow up here as soon as we know more!

It's a fair question - unfortunately not one I have an answer to but rest assured I've delivered your feedback to the team!

Hey Peter, this is clear and I'll pass on to our eng team now!

If you write out the problem and specific feature request I'd be more than happy to file a new feature request on your behalf.

Unfortunately at this time there's not a native way to do that natively. For this kind of automation you may be able to push in 10 rows at a time via webhook orchestrated via some sort of external script or automation tool. I haven't actually seen this in practice, but it may be something to look into. Otherwise manually running in small batches is likely the best way to get around this at the moment. All this being said, the team is aware of this kind of issue and is working towards making the product more robust in this regard. Sorry for the inconvenience and I really hope this helps.

Hey there, could you send over a table link so I can raise this issue to our technical team?

Hey Peter, You can avoid the circular dependency entirely by using “Choose number of rows to run” on the AI column instead of conditional logic. ​

This lets you run the same AI prompt in smaller chunks (e.g. 10–25 rows at a time), which tends to preserve accuracy without adding extra columns or dependencies. You’re not re-running rows or changing credit usage — just controlling how many execute per run. ​

In practice, you’d run the first N rows, let it finish, then run the next N rows. This keeps the table simple and avoids the batch-related issues you’re seeing when running hundreds at once. You can access this option by clicking into the column header, clicking run column and then choosing number of rows to run. Hope this helps :)

What you need to do to avoid this is to extract the linkedin from your enrich company column as its own column, do the same for your claygent column and then use a merge column enrichment to merge those two columns. Once those columns are merged you can use the resulting column as your input for downstream enrichments :)

Hey Yashar, as a first step I recommend using a clean format for the domain like this: https://naviaenergy.com/ or just the isolated domain (naviaenergy.com for example) To do so you can use the normalize a domain enrichment or use a free formula column. If that does not improve results I recommend using a claygent search to return the missing LinkedIn profiles. I hope this helps :)

Hey Christopher, in this case I recommend adding an additional Claygent column to research the lowest level entity. To do so you can use a prompt like this: "Find the specific website domain for [subsidiary name] located in [location]. If they have their own domain separate from the parent company [parent domain], return that unique domain. Otherwise return the parent domain." This should help resolve those lower level domains. Let me know if this is useful or if you still run into issues :)