Request for Custom Key Queuing to Improve AI Enrichment Process | Clay

Request for Custom Key Queuing to Improve AI Enrichment Process

Clay team, please please please (for the love of god) institute queues for custom keys for AI. It takes SO MUCH TIME to rerun enrichments with private keys because there is no queuing for private keys it's a game of waiting until everything errors out and then having to go back.

4 comments

Channeled
APP
·
·
We understand your frustration with running enrichments using private keys. While we don't currently have a queuing system for custom keys, there are a few things you can try to improve your experience: 1. Process data in smaller batches (around 50 rows at a time) to manage API limits better. 2. Check your API tier and limits. You can view and adjust these in your provider's settings (e.g., Anthropic API settings). 3. Consider using our Clay-managed key option for smoother processing, if possible. If you continue to experience issues, please request human assistance by clicking on "Talk to a person" for more tailored support.
Source: Use AI integration overview | Documentation | Clay University
Channeled
APP
·
·
Jorge M.
·
·
I think I'm running into the same issue. Whenever using AI (no Claygent), you need to wait until the runs finish to verify if everything went well. Would the solution for this be to have running schedules with time? For example, I see an anthropic prompt that goes into the web, scrapes something, and returns a yes/no value, which takes around 30 seconds to a minute per row. It would be easier for me to just run 10 rows every 10 minutes in the background when I wake up the next day my table is done vs. having to manually run those myself. (really painful process)
Channeled
APP
·
·
Good news! We've recently introduced a Scheduling feature that addresses your need for automated AI runs. With Scheduled columns, you can automatically re-run specific columns or entire tables on a recurring basis. You can set up your AI enrichments to run daily, weekly, or monthly, keeping your data current without manual updates. This feature works for both sources and columns, allowing for a more efficient workflow. Additionally, you can use the "Auto-update" option in the Run Settings to automatically enrich new rows as they're added. You can also set conditions to control when the AI runs using the "Only run if" option. To get started with Scheduling, check out our detailed guides on scheduled sources and columns. If you need further assistance setting this up, please request human support by clicking on "Talk to a person".
Sources: Use AI integration overview | Documentation | Clay University Build a sustainable growth engine with Scheduling - The GTM with Clay Blog