Conduit University #4 — Inbox

Hi, I'm Emily and I'm an agent product manager at Conduit. Conduit uses contact based inboxes. So, the tools and

based inboxes. So, the tools and workflows are tailored to fit the needs of each specific inbox. Today, we're

going to go over our inbox zero philosophy as well as the tools that you can use in your inbox to reach inbox zero.

The way Conduit's inbox is set up may be different than how you may have been interacting with an inbox. Think of it more like a checklist than a traditional inbox. In Conduit's app, conversations

inbox. In Conduit's app, conversations are going to be in three places. It will

be in to-do, follow-up, and done. You

can also use filters to change how you view your inbox. Conversations in to-do will need attention by either your AI or your team. If you encounter a request

your team. If you encounter a request that you want to action later, either because you don't have enough information or want to wait, then you can use the follow-up feature, which moves the conversation to follow-up.

Finally, once a conversation has been actioned, you can move it to done. If a

customer responds to this message, they'll move back to to-do. Your goal is to get a to-do account of zero by actioning customer requests. Once you're

in your inbox, we provide a variety of tools for you to reach inbox zero and respond to messages. You are able to reply directly to messages, assign them to a team member. You can also leave a

note so that other people can understand.

Conduit is not just an inbox organizer.

It can completely change how your business uses AI and you're able to see how the AI agent thinks and teach directly in your inbox. With Conduit,

every AI message is explainable. You can

see its reasoning, the sources it retrieved, and you can even teach it within the conversation. We want to move your account to autopilot where conversations are being replied to

automatically. But if the AI is not able

automatically. But if the AI is not able to answer a question or it has an urgent tag, then it will not be responded to by the AI. Let's say a customer is unhappy

the AI. Let's say a customer is unhappy or is experiencing an emergency. The AI

will not respond and add an urgent tag and give a proposed reply to the urgent message. If the AI ever needs your help,

message. If the AI ever needs your help, that means it needs human action.

Whereas AI missing info means that the AI does not have enough information and you can add more information to the knowledge base. In your inbox, you're

knowledge base. In your inbox, you're able to look at proposed replies or replies by the AI and see how it analyzed the conversation. on the right.

You can look at the sources it used to get to that response. If you want to teach the AI more information or make edits, you're able to teach within the inbox. There are two ways to teach your

inbox. There are two ways to teach your AI. The first is within agent hub and

AI. The first is within agent hub and you can write out the information you want to teach. The second is when you are in a message in an inbox and you can click K or teach AI. You can also test

the AI in the inbox using the new knowledge that it learned. In the

sandbox, you can simulate the conversation to test the AI and make sure that your new knowledge is working correctly. Once you understand how to

correctly. Once you understand how to use these tools, you'll be able to collaborate directly with your AI and reach inbox zero.