Building with AI Agent Teams

Like human teams, AI teams bring together specialised agents that collaborate on projects and complex tasks.
AI teams have one manager that overseas a group of AI workers.
This manager agent hands out tasks to team members based on what they’re best at, and makes sure all agents have the information they need to succeed.
Here’s a closer look at how to think about and build with AI teams.

[@thewiredworld4791]

How do we KNOW the SEO is optimized?

[@JorritvanGinkel]

Seems really cool, can you now also chat with the manager to give feedback on specific things and improve them?

[@relevanceai]

You can yes! We've built out other simple but effective examples of interactions between blog writing agents and fact checking agents who give feedback and request changes that you can view on our Linkedin.

[@NinjaTaskers]

@@relevanceai where can I watch about this example?

[@AIPlayGrounds]

When will be available for everyone?

[@presonator]

I really can appreciate the enormous significance and potential of this sort of approach. My problem is that despite enthusiasm - I am never quite as adept at learning new tools such as Relevance. I'm wondering though if I'd be better off waiting ( a few months?) for Relevance to add a shopping mall where I can add various roles to my cart, throw in a manager and checkout with a whole team. Is that where we're heading? Cheers. Rob

[@theetravellover]

Are you referring to whitelabeling?

[@presonator]

@@theetravellover Not really - that implies that I want to resell agents to others - On reflection, I suppose I meant to say I look forward to continued refinement in the interface/drag n drop functionality and probably most importantly tutorials - so that tentative users feel more confident to build stuff. 🙂

[@relevanceai]

We have a lot of tutorial videos coming out soon!

[@presonator]

@@relevanceai - Super (Supa?) excited to see more of the Supademo tutorials - they are fabulous! Thanks so much

[@PatrickSteil]

Question. It seems that everything you did here with the various agents could have been done with a “tool” for each one, no?
What is the advantage of using an “agent” vs tool in this scenario?
Thanks!

[@MuseStorytelling]

also wondering the same thing here...