Gather extra info
Advanced AI-Driven Workflows for Complex Support Needs
đ€ Your AI Assistant Canât Read MindsâHereâs How to Help It Dig Deeper!
Not every customer query comes with enough info to solve the issue upfront. Thatâs why your AI needs a smart system to ask the right follow-up questionsâand thatâs exactly what weâll cover in this lesson.
Youâll learn how to use intents and missions to collect missing details, guide the conversation, and give your AI assistant a second shot at solving the issueâwithout instantly escalating to support.
đ What youâll learn in this lesson:
- đ§ Define specific intents like âInvoice Sending Issueâ
- đŹ Set up missions to gather more context (like missing emails or error messages)
- đ Let Aura loop back with the AI assistant once it has enough data
- đ§ Troubleshoot with smarter, context-rich AI responses
- đ§ Reduce manual workload and limit unnecessary escalations
đ Real Example: A user says, âI canât send my invoice.â Instead of guessing or escalating, Aura asks targeted questionsâdetects missing dataâand then hands it back to the AI assistant with the full context to resolve the issue effectively. đĄ
Transcript
Auto-generated from the video captions; it may contain errors.
Now sometimes you got questions that can lead to a back and forth to collect more data, more information about what the customer is experiencing and then after collecting the data it can lead to a simple answer really and so your AI assistant will be able to assist in that case but with the very first uh iteration the very first question your customer should it wasn't possible to properly respond to the query. All right. So, in this lesson, we're going to see how to create an intent, ask for more information about that query, and then pass it to the AI assistant again to try to find an answer.
Okay. Okay. Let's get started. So, first we're going to define an intent. Okay. So, the intent would be something that is related to the issue. uh the broader issue your customer is having. Um so for example it might be related to the sending of the invoice. Okay. So invoice sending issue. A user is reporting issue regarding the sending of the invoice. This might happen for different reasons right? So it might happen because I don't know the user email is not there. the invoice is not being finalized, you know, something that requires more back and forth. But sometimes the user just don't know how to ask something that specific and they just ask um I don't know, I I cannot send my invoice anymore.
Now that you define the intent, you can uh add a mission over here. And the mission it would be prompted with a goal that is to better uh help the customer uh ask them to provide more information on what's happening with the invoice. uh if they see any error um if they um completed the uh invoice. Okay. So this would be a general question that aura will ask uh after um the user reported the invoicing issue. So this after there was a success it will be easily handled by the AI assistant again and if something goes wrong I can now assign this to of course the customer support team.
So here we go and of course I will be able to close the conversation over here whenever the assistant uh solve the issue after iterating or if there is an escalation again. Uh this time I'm going to assign it to support because maybe it requires more information. Okay, here we go. We are going to save this as invoice sending issue and we're going to publish it. Now let's test this out. Now I'm reaching out to the software team saying I cannot send my invoice just straightforward. you know this might happen. So right now Aura instead of searching an answer in the knowledge base is going to match the intent and so she's going to gather more information.
She's going to try to get more information about it. I see an error of um insert not updated. So now Aura is again passing this information to the the assistant. So let's go under your inboxes and see what happened. Um again it seems like she's still in the mission mode and she's still trying to gather more context about this. So thank you for information. Can you pro provide more details about the any additional context or specific error? Um, so this is the only error I've got. Um, I'm sending this to a new client, Emma. So now what is what she's doing after collecting all the data, she's already handing off this to the assistant.
And as you can see now that she gathered new information, she went into assist mode. Okay. So, she is now trying to provide more uh more help based on your knowledge based articles, your your candidate responses. This is the way you can dig deeper into a single customer queries that sometimes is so vagnot in one single iteration. This is a mix of um using the AI. This way you will be able to gather more information. Aura will try to solve it and then pass it back to the AI assistant. The best way would be like for example if you do have several issues that can arise whenever they're sending the emails such as I don't know with a new customer you need to first save the customer or something like that then it would be great to have some troubleshooting articles for each of those cases or like just
one troubleshooting article with all of them and then so the context would be there to respond to the various specific case but in this way aura um basically asked more information about that issue and then provided more guidance to go through the troubleshooting. Okay. So this is an advanced way of using your AI assistant, the integration, the missions and again the AI assistant to try and find a solution which will lead to less back and forth between your customers and your support team. Thank you so much for watching this video. I hope you like this lesson. In the next video, you're going to be assigned a great task to explore some advanced automation with your AI.
Thank you so much for watching. Ciao.