How You Can QA Contact Center Calls Using ChatGPT (Desktop)

00:30advice from a call center geek, a weekly podcast to focus on all things call center. We'll cover it all from. From call center operations, hiring, a culture technology and education. We're here to give you actionable To improve the quality of your customer's experience, this is an evolving industry with creative minds and ambitious people. Like this guy. Not only is his passion call center operations, but he's our host, he's the CEO of Expivia Interaction Marketing Group, and the call center geek himself, Tom Laird. the desktop version just because that's where most where most everybody is at. [pricing remark omitted]

[Gap 01:26–01:56: no reliable transcription]

01:56and I just didn't at the time time to play a year and a half You know, kind of the learning more about prompting, right? I mean, promptings only been around for about a year, year and a half, right. So we went to school on, on that and what are the what are some of the things that that help tobraska make things is more consistent and make things better. It kind of gives the instructions, and then we got into the asking questions and prompting for those questions. So there's really four main pieces to this, the system prompt, there's your questions that that you're asking, we call it the context, but it's basically the prompt you know, and I can kind of give you this thinking this is pretty basic, you know, you are the head of quality assurance for a contact center, you oversee answering questions, you will examine the attached transcript of the answer, right? Everybody can figure that not proprietary anything, but then you start to think the call a agent-sentiment and some of the things that we're doing and are prompt for this in our system prompt.

04:22This involves a nuanced examination of language, tone, and the context presented in the dialogue dialogue, pay close attention to key indicators such as word choice, intensity of expressions, and just define sentiment. Next prompt is we want to do sentiment score, right?, for this call that we to say, hey, this is what customer sentiment is, this is what I think is really cool, we call it the

[Gap 05:44–06:14: no reliable transcription]

06:14and it'd say, no, I think, I think that's really, really important stuff that we have found to be really important in the output. The other thing that I think is really important too is, is again, if you don't have any programming Just certain things that you can ask for, right?

[Gap 06:29–06:59: no reliable transcription]

06:59and struggled with and then everybody's paying more for auto summarization right ask it to cost

[Gap 07:05–07:14: no reliable transcription]

07:14you can use a personal GP

[Gap 07:16–07:44: no reliable transcription]

07:44this up is to tell chatgpt, you're on your desktop, um, If a call is answered with a yes, a the question is answered with a yes or five points.,,,,, out what we wanted to do from an output standpoint, how consistent that. We wanted it to be getting the real outputs, how, how we wanted to leave look on a form and an evaluationincerity and then we get to the questions right and that's where the rest of the you know seven eight months of. Work have come in the throat for the lawyers don't want to totally have me Have me stuff is very basic right from a

[Gap 09:08–09:38: no reliable transcription]

09:38And you trial and just, copy and paste it right into your chat to the. Instance, right and then hit enter. If your output isn't in the format that you like, then go tweet that prompt. But once you get it to where you, like again, I think that's where the cool personal G.P.T. CPTs can really come into play for some basic stuff to where you can basically upload especially your system prompt and then maybe Could we build the question and we build all, all this stuff out before the person and the answer was yes, so that was K that was our proof of concept. Some tips that I have for you on prompting. We, have found that imperative is a very. Strong word. So if were you say it for smaller type prompts, having, having a paragraph is fine, right? But if you get to a longer prompt, the order having it almost a in a checklist, definitely gave us more consistent and gives us much better results.

12:11The other thing is, the models do a great job with keywords, finding specific things, but don't think that way, think much bigger. If the So those are some things that I,.

[Gap 12:37–13:05: no reliable transcription]

13:05Right. Right. So when I, you know, you know, all auto QAs are doing is writing prompts. That's true. I'm. I. In. In. In. In. In. Or it You can upload more than five. Say you upload 15 chat transcripts, call transcripts, and start querying. What percentage of the customers call were irritated. How many specific customers And you can start to kind of build that, see how many you can get, and start to learn more off of your data as well. You can do a ton of that stuff with your personal view of uploading a ton of different data, and then querying off that data. So there's so, so much you can do internally to build some things out yourself, even a very team, team can kind of can, can do some really cool things with the technology of today. Again, again if you're that, if you're that person that just wants to kind of play, play around and figure-out some really cool ways to just start to score calls. And I think, and bridges to other other opportunities within the contact center as well because and then take it to the next step is get your personal GPTs involved in a little a a little bit more at scale. If you want to see what you're doing at totally at scale with with API connectivity built to get from an audio standpoint and also to multiple large language you'll be more than happy to kind of show you that and how that works and then you know using on top of that, right? It's a process to do, but I think, but at its core, you know, anybody can start to play with this and start to get some results, at least on a smaller scale for calls. So, if there is no questions on that, you know, if you guys are listening to this, please DM me if you have anything that you want specific answers to, be more than happy to help, but I hope that was kind of cool and kind of gives you a little bit of insight into how this works.