Bernard Slowey

And which is like you have data that, you know, content is a good example where maybe you have a couple of articles on something that's similar and maybe you're using like an acronym name or something. And the agent can pick the right piece of content. It's a collision of that content. And so so that comes back to this kind of data hygiene piece that you asked about earlier.

It is really important that you have the right data. You're you know, we've spent a lot of time actually calling older articles. stuff that's not relevant anymore so that the agent's not trained on it. So that's definitely a very, very common one. The other one I would say is the experience from agent to human. We talk about that a lot. Because in our experience, it's like you have a choice.

It is really important that you have the right data. You're you know, we've spent a lot of time actually calling older articles. stuff that's not relevant anymore so that the agent's not trained on it. So that's definitely a very, very common one. The other one I would say is the experience from agent to human. We talk about that a lot. Because in our experience, it's like you have a choice.

It is really important that you have the right data. You're you know, we've spent a lot of time actually calling older articles. stuff that's not relevant anymore so that the agent's not trained on it. So that's definitely a very, very common one. The other one I would say is the experience from agent to human. We talk about that a lot. Because in our experience, it's like you have a choice.

You can have the human join the chat And so, you know, they come in, it's a seamless experience or maybe you don't have the time at that moment. So you can create a case to like have someone come back to you at a time that works for you. And so we spend a lot of time designing that experience. And so a lot of customers are like, how did you do that? Show me how it works.

You can have the human join the chat And so, you know, they come in, it's a seamless experience or maybe you don't have the time at that moment. So you can create a case to like have someone come back to you at a time that works for you. And so we spend a lot of time designing that experience. And so a lot of customers are like, how did you do that? Show me how it works.

You can have the human join the chat And so, you know, they come in, it's a seamless experience or maybe you don't have the time at that moment. So you can create a case to like have someone come back to you at a time that works for you. And so we spend a lot of time designing that experience. And so a lot of customers are like, how did you do that? Show me how it works.

And so that's another one that we spend a lot of time on as well.

And so that's another one that we spend a lot of time on as well.

And so that's another one that we spend a lot of time on as well.

I can't remember the name of the bias, but as humans, we have a bias for machines that machines should always be correct. And the example, the great example is like self-driving cars, right? Like we're afraid to let them onto motorways because there's like 0.000 whatever percent chances that we'll have an accident. Humans have accidents all the time. Right.

I can't remember the name of the bias, but as humans, we have a bias for machines that machines should always be correct. And the example, the great example is like self-driving cars, right? Like we're afraid to let them onto motorways because there's like 0.000 whatever percent chances that we'll have an accident. Humans have accidents all the time. Right.

I can't remember the name of the bias, but as humans, we have a bias for machines that machines should always be correct. And the example, the great example is like self-driving cars, right? Like we're afraid to let them onto motorways because there's like 0.000 whatever percent chances that we'll have an accident. Humans have accidents all the time. Right.

So self-driving cars are actually a safer option than a human. Right. But we don't want that to happen until it's near perfect. Right. And so it's similar with some of these agents. It's like people expect the answer to be perfect all the time. And it might not be. And it might not be for multiple reasons. Like we talked about content.

So self-driving cars are actually a safer option than a human. Right. But we don't want that to happen until it's near perfect. Right. And so it's similar with some of these agents. It's like people expect the answer to be perfect all the time. And it might not be. And it might not be for multiple reasons. Like we talked about content.

So self-driving cars are actually a safer option than a human. Right. But we don't want that to happen until it's near perfect. Right. And so it's similar with some of these agents. It's like people expect the answer to be perfect all the time. And it might not be. And it might not be for multiple reasons. Like we talked about content.

If the agent doesn't have access to the right content, if that content's wrong, then your agent is going to be wrong. Right. You know, and so so there's so many reasons, like there's so many reasons behind it. Like I gave you the example earlier on of like, if you don't like the prompts are in position the right way, then it might respond in a different way than you want it to.

If the agent doesn't have access to the right content, if that content's wrong, then your agent is going to be wrong. Right. You know, and so so there's so many reasons, like there's so many reasons behind it. Like I gave you the example earlier on of like, if you don't like the prompts are in position the right way, then it might respond in a different way than you want it to.

If the agent doesn't have access to the right content, if that content's wrong, then your agent is going to be wrong. Right. You know, and so so there's so many reasons, like there's so many reasons behind it. Like I gave you the example earlier on of like, if you don't like the prompts are in position the right way, then it might respond in a different way than you want it to.

And so there is that expectation setting. And that's why I think that experience agent to human is critical because you need to make it easy that someone can go, oh, actually, you know what? I just want to go true to the human.