Update Answerly

@Simone @Fatos @Nevrie

Untill your update, we had 1 agent attached to 1 knowledgehub / training collection.

Could you please explain with a use case why the below is an advantage?

Now, you don’t need to train agents individually. Instead, create a training collection and associate it with each agent as knowledge/memory.

  • Do your mean that when we have 1 agent, we can swiftly switch between collections to see what works best?
  • Why would I create 2 different agents, using the same collection? What is the use case ?

Or what did you have in mind with this?

Thank you for you explanation.

Good questions.

Hey there! :wave:

Here are the answers to your questions:

  1. Centralized Training Collections: Instead of training each agent separately, you now create a single training collection that can be linked to multiple agents. This means if you need to update or refine the knowledge base, you only have to do it once. All agents tied to that collection will immediately benefit from the update. It’s a big time-saver and ensures consistency across your team.
  2. Flexibility for Optimization: Exactly, with this setup, you can easily switch between different training collections to see which one performs best. This flexibility allows you to experiment and fine-tune your agents’ responses to match real-world needs more effectively.
  3. Different Agents, Same Collection: Why have multiple agents with the same collection? Imagine you have different teams or departments that all need access to the same core information but with slight tweaks to their responses. For example, your support team might need detailed technical answers, while your sales team focuses on product benefits. Using the same collection but creating tailored agents helps meet these specific needs while keeping your core knowledge consistent.

In a nutshell, this new approach makes managing your training materials much simpler and gives you the flexibility to adapt and improve your agents as needed.

Hope this clears things up! If you have more questions or need further examples, just let me know.

1 Like