There’s some bugs with your more advanced models and it performs worst as Claygent that the cheaper models I made a screen recording https://screen.studio/share/sEEPbpet?state=uploading Ultimately I’m trying to find
How many people are in the “IT departments”
How many are “senior”
LI URLs of the people found
Maybe my approach is not right but nonetheless thought it worth sharing that the better models fail worst than the weaker ones…
Hi there, thanks so much for sharing this! We’ll be passing your feedback along to our product team, and any additional context you can provide will help us prioritize with your needs in mind as we shape the roadmap.
Want to chat with someone on the team? Feel free to hit the button below!
That’s actually super valuable feedback thanks for sharing. You’re right that sometimes the higher-tier models (like GPT-4o) inside Claygent can overthink or get rate-limited on structured queries, especially when you’re asking for counted results + LinkedIn URLs in one step. The cheaper/faster models (like Claude Haiku or GPT-4o mini) tend to perform better because they parse the results directly rather than reasoning through summaries. For your specific use case (IT department size, seniority, and LI URLs), a more reliable approach is:
Use “Find People by Department” to isolate IT.
Add a formula to count and filter for senior titles.
Pull LinkedIn URLs directly from the enrichment instead of relying on the AI parsing.
You can still use Claygent to summarize insights (like % seniority), but let structured enrichments handle the heavy lifting that combo tends to be both cheaper and more accurate.
Find People by Department is an enrichment per se or you’d use claygent for this?
I tried this way and it’s failing It stops at 2-3-4 people when there could be 10-20
Ruben clicked 'Talk with the team!'
Clay Support is sharing a file with you.
Our support team has got your message and we'll get back to you soon! We're currently outside of our standard office hours (9 AM to 9 PM EST, Monday through Friday), so it may take a little longer for us to respond.
If you’re dealing with a specific table, drop the URL below so we can help you quicker. Otherwise, someone from our team will be in touch soon!
Hey Ruben - thank you for reaching out. Could you send the table URL so I can take a closer look?
Hey Ruben- jumping in for Ashwin here. Could you send me your table link and I could take a look?
.png)