Hey community 
We’re working on a new Evaluation feature for AI workflows.
As you know, AI workflows can be unpredictable, so we’re building a way to evaluate them properly and make sure they’re reliable and return the expected outputs, even when you tweak a prompt, switch models, or something changes under the hood.
We’re looking for users who have experience building AI workflows and are up for joining a 30-minute usability testing session to give us feedback.
If that sounds like you, drop me a DM. Would love to have you involved!
Thanks 
77 Likes
Would love to be involved in this - I’ll drop a DM
I am interested, you can call me via email: [email protected]
We love to join this event, please mail us [email protected]
1 Like
I would love to join in this event, send me a DM.
Dunno how to DM you but I can help. [email protected]
I’m showing my age here because I can’t figure out a way to DM you. but I’d be very interested as human in the loop would make a huge difference to the reliability of AI workflows as you’ve pointed out. Can this run on the self-hosted version (the test, that is)?
If so, you can reach me at: [email protected]
would love to be involved, we have built and cursed 
a lot of Agentic Workflows with a ‘manager’ agent.
I have a few workflows using AI nodes and am actively developing new workflows with AI.
I am not a developer but more a hobbyist. I would like to try it out and get involved. I tried locally n8n on my nvidia AGX Orin in the past.
I’m interested. Call me if necessary via DM.
hey @giulioandreini - count me in! 
We have started to build an integration, would love to discuss [email protected]
Hey Giulio, I’d love to help out with this. Count me in!
