How to Automate Your Contact Center QA with Ai!

00:00Hi guys, Tom Laird here. You've call center geek, a call center geek, a call center geek the focus on all call center operation, hiring, culture, technology, and education. We're here to give you items to improve the quality of your customer's experience. This is an evolving industry with creative minds like this guy. Not only is his passion call center operations, but he's our host. He's the CEO of Expivia and the call center geek himself, Tom Lier. Welcome back everybody to another episode of advice from a call center geek Beautiful sunny and kind right now, how was everybody doing? I want to kind of talk. I think that And I think that this is a little bit of a special episode, a lot of questions that have the post that I did last week on LinkedIn on how we've we've been able to somewhat to the point of being able to totally automate QA or at least have the proof to automate our quality services in our contact center with all of forms that our client utilizes our clients you know. This was kind of an eye-opening experience for me. So I don., have all the answers on this. I'm not an expert on this. We are not a software company. This is something that we've tried to I think as it may kind of come out, to be and to be honest I think internal content and non-third parties, right I think it's almost.

02:54software, not to say that there's not stuff out there now, because of course If on tik-tok as well. I'm gonna do this as an AMA again. Any questions, any questions that you want to talk about with this stuff, I'm going to get into it. How we've done it how we're doing it and let me know if there's anything that I can kind of help. Again I'm an open open book with this stuff. It, I think this is going to be a product that I think that we can really go sell because to be honest I think it's too easy.

[Gap 03:52–04:22: no reliable transcription]

04:22Trump NPS. [QA coverage remark omitted] do, exactly like an agent. Now these are in-depth prompts, right the difficulty, it's really easy to do if you can figure out the prompt. It's very difficult to do if you if. If you can't and you don't understand the prompt, or you're in some of that manner, Which we've done a lot of experience with to kind of figure that out. So the first thing that we did is, is we said all right, we need transcripts. I

[Gap 05:35–06:01: no reliable transcription]

06:01CH CH Again, I think one of the added benefits of having analytics is you do get the transcripts. And we are still working. So let me again, let me say this as well. This This is a work you started to test things out test our prompts out and some of these can you that you s a 90 I think that was another problem right we're automating it with a yes, yes or a, the the

[Gap 08:13–08:31: no reliable transcription]

08:31or if they had a lower score, would be extreme to do in a time consuming way for for an agent. Taking, writing prompts easy questions. That's an easy one. What happens if this is I know a lot of you guys, right? And let me just a this to the two things that you need. I, I mean, I think that again, I think all this stuff is obvious, but it's not being said. The two things, things that you don't want is we need to be able to have access to to the data, right? So,., whatever model that. well. So there could be four or five different types of calls. somebody canceling right in all of those types of calls correlate to different questions that are being asked right everything from verification to asking you know what actually happened on the call to disclosures at the end to making sure that everything from the compliance issue if we do have a sale from masking, doing all the stuff, that becomes extremely more robust.

10:29I and I don't want to say DAPA but I think more robust when you're thinking through the prompt, right. So you have to have a really analytical, logical mind. Because again, this is still a kind of computer process, but you have to have the program knowledge and kind of the art right with that too. So again, we've been we're trying to bring in our whole team of I T of management of the on floor. Guys that deal with this every single day are obviously are to come a lot of if-sans that a part of, specific types of a specific type of call that would then move down the and to a specific type M of questions from a Q.A. from a Q And what does it I mean, and to be honest, you know, we have seen slight changes in verbiage, right, do have some some significant in what happens with the, the, the output of the, the score.

12:08But I think when we got it to a point where we felt comfortable and then we will correlate We would correlate We, we would correlate it with our, our QA. We call them EQA, Exp. I think, from that aspect, we were doing, you know, a pretty good job when it came to how, how do we now scale this? Because we feel like we have a Yes, I mean, I'm not there yet, to be totally frank. We're not there yet. We have connectivity. Our APIs are connected to ChatGPT. We have been. We've tried to scale and when we try to do 10 to 15 different of the same thing. We're looking at that today. different calls and everything looks pretty good-good, so we need the tweak there, but from a cost standpoint, you know, what we have found is ChatGPT or 5.5 we can We, can get pretty close to the same answers that we're getting in chat g.

[Gap 13:30–13:56: no reliable transcription]

13:56with the input and the output cost using the API. With ChatGPT 3 or 3.5, to work with that, it was great. It was thousands of calls. That's something that we have to look at. [QA coverage remark omitted] But. But we can talk about how to do that as well But I think you know from a value add there that makes total sense.

[Gap 15:06–15:15: no reliable transcription]

15:15have to 200 programmers just working in Foc S. S. and I think that there will be a human toll, when it comes to scoring. the I think that this a low hanging fruit where we could see AI really take some take, some of and go and start really coaching with that. So again, from a cost standpoint, I think it is negligible when the third part when the C cast players come out, I a thing is going to be exposed. So, I mean, I think it, and almost paid through yourself. Now, granted, it's not Not there yet, there's HIPAA issues, but how far this is right now is pretty insane. What you can do right now is crazy, and the scoring scoring of calls at scale if you have a good prompt is really amazing. All right, another question. There's a lot that can be done in the- space, using track PT is I've just I think that that.- that's what is going to happen eventually, right?

16:57That I think people are going to- almost have their own instance of ChatGPT,. Like very large companies are gonna do that. I don't care about the learning point and and I think that's where we get we go too far right the the AI learning of what you want to do is something that's going to happen and it I mean, I mean, it'd I'd be awesome, but we're just kind of taking a kind of a tool that we're doing that probably doesn't have to be done by humans, right, and we're scoring calls and then giving kind of just some insights into what happened on that call and how to get things done. So from a learning and a machine learning, I don't really care, to be honest, to be honest, I mean, I don't think to, that, that's like a roadmap cool. as close to a human being.

18:11Okay, I think what we've done is cool, I think, maybe just only we talking about it. I am positive there's a ton of you guys out there or some of you guys out there that that I've already kind of figured this out, right? If you can program and you have the logic to program, Then you definitely have the brain and the logic to develop specific prompts on different types of audits, on different types of QA. You can score calls exactly how a human being is doing, asking the same type of questions that humans are asking, but you're actually just typing it out and having the ChatGPT process do it. Now, there's security issues right now. Now, the good thing is, It made me feel at least

[Gap 19:15–19:34: no reliable transcription]

19:34from a HIPAA standpoint, obviously, there's some things that need to be pre-done on the script. And I think that's an AI solution. There's something something there as well. But yeah, guys, I mean, literally, literally, I cannot explain that in 20 minutes. That's really all we're taking our chat transcripts from analytics. We're pulling them in we're them in. We're indexing them so that we know, you know, who was what and what program We're indexing them, so that we know who the outputs, in a cleaner, better way., there's no APIs right now that we have with nice CXone to put them back on the QMA platform, which would probably happen. If we can figure out the scoring, we'll be able to put them back in so that we can be able to maintain. I think that this will be an added, a huge benefit for our customers. I think, I think that if program or on your staff, you can start to figure this out to do on your own.

21:00Don't get too complicated with it. embattled start thinking like we have to do all of this like crazy machine learning. You're just basically basically having a it's the. The same thing of transcribing a call at the and putting it into a CR a CR. There's no learning that has to happen with that. It's a job that it needs to handle. You don't get caught in the weeds then with a lot of the things that people are thinking through. Take that first step. Develop your prompt. Practice it on the UIUX, on the ChatGPT-4, or. Find a prompt that works. find a prompt that works right go to your QA have them score the calls I'm okay with that right We which I know can be that, to be honest, that's the hardest piece, the it's not that, it's not that difficult. And while there still needs work to be done on this, think about where this thing's going to be in six months, just six months.

23:18You know, huge players that are working on this stuff right now. And I think it'd be very difficult. I think you have to almost have an onboarding team to sit down and really go through, you know, like Kapil Sharma here on LinkedIn, you know, he has a really good point. They have customers with 30 plus different types of audits. And there's so many paths that QA can take based on human interaction, what humans can say. But you have to be able to account for as much of that as you possibly can with the prompt. But if there's a hand holding to a prompt that everybody loves and likes, and then they can just pass it, this thing could be done. This thing could be done at scale. this thing can be done as a business. I know a lot of the CCaaS players are kind of thinking that through, but from what I'm hearing, there's more arbitrary just kind of numbers instead of really looking at like a BPO method of, we have 25 different clients and everybody has their own platform, has their own form.

24:22How do we score that and kind of look through that? So yeah, guys, that's kind of what I got. I mean, I don't know if there's any more questions on there. Again, it's really cool, but it's really not that difficult, right? There's some things in it that are maybe unique to kind of get through from figuring out the prompt, from figuring out how to get the transcripts, from what do you do with the output, but that's kind of the fun part, right? Especially the output and kind of figuring out the prompt. Get your team together, have some fun with it. I think that you can really do some from that you can practicing and testing it on your prompt

[Gap 25:15–25:28: no reliable transcription]

25:28can hammer some type of low, the low hanging fruit, which is, you know, the scoring of calls, right, it's just like, like, research, it's like this, you know, just give a good data in and it's good data out doesn't need to learn doesn't need to get crazy into certain things. is. I'm trying to.I.P.T all the time not just for like cheating on school right but but understanding

[Gap 26:37–27:37: no reliable transcription]

27:37is the CX Rockstar if you've ever seen him at SCCW or any of the big- I mean the big contact center shows he's a kind of the Rockstar the the Rockstar CX in the AI's as it relates to the contact center so we're gonna have a you know one hour-hour AMA just talking about the contact center AI all a new stuff But that's kind of coming Let's called a little love here boys and girls. Where you from man, so I'm from Buffalo, New York. Oh, nice. I wrote in an eerie Pennsylvania. I love going to the Mill Creek Mall.