Dear Santa: Here is the AI Contact Center Tools We Actually Want!

00:00Tom Welcome back everybody to another episode of the McCall Center Geek The

[Gap 00:32–01:00: no reliable transcription]

01:00fully score and give analytics and insights into all of your contact center calls. I know I've been saying this same kind of thing in the last probably four, four or five episodes, but, you know, there's been going on here when it comes to again, just personal life. Like I coach high school basketball,

[Gap 01:18–01:36: no reliable transcription]

01:36you have questions about the platform, and again, I've done a newsletter on this, I've done multiple I have multiple posts on this, but this is kind of the, I guess, I guess maybe the most robust conversation I want to have on. We're just the CCaaS platforms, these technology, major technology partners, they're building the. The same stuff and it's. Kind of boring to be honest, right? There's a race to I guess, you know, use our agent to take, over your customer support. Like that's the race. That's the only race that people are running when there's a you're going to do a video at the end of the year on what I see as the real much deeper level, but these are the over your support and give a customer a another option for self-service when we move to it, then that those are I believe the companies that are really going to thrive.

04:23And if it is or the Genesys or the talk desks are the Five9 or the you jets or any

[Gap 04:29–04:39: no reliable transcription]

04:39I don't care what anybody says. It's not easy easy to use. AI is not easy to implement. Not yet. In 2025, 2026, when you're using these large platforms, it's very doable. It's not like it's brain surgery, but it's not a just set it and forget it kind of thing. There's a lot that goes into But anyway, so here's here's I don't, five, six, seven depending on how deep I want to get into this of things that I would like to see. Seecast. This would be my AI Christmas wish list right for the Seecast companies. would just, it would, it's probably not being bilt because it would put a lot. lot of people out of business. But I don't think it's that difficult to do when you have the resources of some of these large Seacaster. Number one is, is kind of an AI flow or routing builder.

05:30That you could prompt in English and really deploy quickly, meaning. Well, every platform has their own kind of programming language or it that you're routing and hey, I want this to route three times, then go to this agent, then look for these hours of operation, and they go all of that kind of stuff. And it can get extremely complex depending on when, when you go to, you know, self-service to an agent, then you're looking and you want to send it to specific agents based on certain criteria. You can round off CRM data, like it gets robust. Normal people can't use the plot to the fullest extent. So there's There's no reason, there can't be an AI agent or some type of AI tool inside of studio. And again, I'm not picking on nice because no one has it, but I just know nice the cornerstone, b.

06:49Right inside of nice that you could say, Hey, build me a script that checks if the callers repeat the customer within 30 days. It sends returning customers, to the same group that they just interacted with. Maybe it, it, it prioritizes kind of the a top tier, the gold tier members., it resends password resets to self-service and you know, escalates any call that exceeds 90 seconds in queue to maybe a your retention specialist or some type of someone who's going to handle a more difficult caller. You, you know the. All of these things, I want to just be able to tell it to do that, and then have it build it out for you, right? Just like a cloud code, if you need some type of API, connectivity or API endpoint, it asks shoot for it tells you where to go. right? Managed services doesn't have to happen, like I'm in managed services I would still prefer to see something like this, right? This would take a massive swath of people probably in companies to doing other things, but this is the probably the number one thing that I think I. I would love to see built that would allow access to your platform to everyone, right?. You don't need to be a huge IT technical person. A lot of these smaller context context, working with, you JPM they don't have this, right?

08:17They don't have that real technical person, so they never use the platform to any extent. There. There's no reason that this, this can't be done. I'd love to see that, but to be a bill, right? And I just, you know, the real opportunity we, everybody keeps still thinking is the, the, another chat bot, but something that builds routing intelligence with English, I think is just vastly more important and more valuable in 2025, 2020, 2026, you know, CCaaS. Down the road, maybe not. But the second thing is something that I've been harping on for a really long time. And we have it in a small kind of micro way, right? Which is, you know, kind of an AI work for, and I hate the word orchestrator, but I'm going to use it, an AI workforce orchestrator that just goes beyond forecast. Like having a mission control, having AI actually run your call center.

09:12You set the business rules, right? And in English you say, we need to keep service levels above 85% for our sales skills, um, 80% for service queues and 75% for general questions. You don't maintain an ASA of 30 seconds for this. Uh, we need to have sales conversions here. Please product as call to our highest converting, um, agents right keep occupancy below or above 72% for the call center. Having all of these business rules set and then having the AI actually move agents inside out, use virtual agents when they have to. No more of this just having somebody staring at a dashboard, moving agents in and out when they start to see queues. We can start to do, and we're looking now at predicted queue times, but adding in your workforce management, seeing when queues are probably going to happen, be able to be preemptive agents to bring them on, that 30 seconds or two minutes before we think that we're going to have a predictive way queue, looking in the IVR for how many calls are in there, moving people before the actual service level takes a beating. All of that stuff, I think, is vastly more important and where we are as a contact center as a whole. Even people that

[Gap 10:46–10:57: no reliable transcription]

10:57. Then lower their priority levels so they don't take as many calls. I get an the email if service level goes under 62%. Please move reserve agents in if the predicted wait time is gonna go under, or the predicted SLA is gonna go under 75. Like we could do this in kind of many ways, and you can, I'd love to see somebody actually build out the tool that connects WFM, connects your staffing, connects your ACD, right? And able to then manipulate and take all of that data, learn from it, and then staff, your contact center, appropriately to make sure that we're hitting certain SLA service levels, ASAs. I mean, even getting into QA and looking at what agents a doing what, right, to prioritize a certain specific agents for certain things, that goes into the studio, I don't even know what the how that means. I mean, you have some AI chatbots. That means you're using I mean I do it too, that's what we say, and that's what we're doing, but I don't have this, this is They have a WFI type tool, if you're not using it, it's normally totally free, make sure that that you ask about that and you start using because we use it all the time.

12:39So I'd love to see this kind of like a WFI on steroids.

[Gap 12:41–13:05: no reliable transcription]

13:05There was a whole thing on really using personalities to route calls with that. Never really never really came into fruition. But I think the ideas there is good, right? But we need to make that next available agent, that type of process of process. We can get a predictive routing that's not procedural. I can use sentiment and tone to understand emotion of that customer. No agent and empathy. We can match customers to agents that deliver the best results for that type of personality. We can look at the notes and see if somebody's been irate, we can look and see When you add in that routing aspect, getting performance that day, looking at their metrics for that day, right? to make sure that we're trying to get the agents to the right place. Cooper, thanks, man. The background is the new ChatGPT 5.2, their new image thing, which is pretty cool.

15:35It's much, much see, and we have started to build an agent for this just internally, it would never be able to be sold, but to have AI dashboards, right, or dashboards for our contact center built by prompt, so right now everything is very static, and I think you have most So So if you have a dashboard, you have your SLA, you have your average speed of answer, you management and kind of your, your, your historical trends, like you, you have all that in a dashboard, but you know, I, I hate to, you know, have to be able to pull We're going to talk a little bit more about that. We're going to talk a little bit about that. We're going to a little bit Again, in a BPO world a customers, and they have these asperger's too, but they'll ask for specific question. Like I, I want to just be like, hey, you know what, show me the handle time by sentiment over the past seven days, compared to conversion rate, right?

16:43Add in a call ball. in service level by skill. drill just my sales team, show it by the hour, highlight where CSAT has dropped more than five percent week over week. What skills does that happen to? And then we get to the point where it's proactive, where it starts to tell you some of the issues that are happening or some of the good things that are happening as well. Again. Again from not just having a static thing that we have to look at and make our Hey, the service level of yesterday in this, our occupancy was here, you know, there's a lot that we could do I think with that. And then it ties back into our, our, our conduct center, right? Hey, I'm going to make these changes now. Should I have them make that change? Right?

17:49AI powered conduct, don't exist, don't exist. So when everybody, everybody says they're an AI powered, AI driven, it's all garbage, right? All they're saying. They use some chat bots, they use agent assist, they probably have auto QA. I say we are an AI powered contact center. That's what we're doing. This again is truly what I think contact center managers would want to see. People who run contact centers, s they want AI to make their life like hands off. That's the whole point of this thing. What we do is we just make AI more difficult. I think that would be very cool and again to have it start to prompt You have to look at different And then it's going to to a little bit more, right? So, and say, you know, you're speaking to fast for this customer. the CSAT is dropping right now, make sure that you acknowledge before you you solve this problem. Your empathy improved 12% yesterday. Your sentiment went up to blah, you're doing a great, you're doing a great job with this and again, I'm not in the ear because

[Gap 19:51–20:04: no reliable transcription]

20:04business rules of the company, so this can can't be just generic. If you're a sales company company, we're. We're going to focus more on the If you're a customer, you're doing great. right now, just as kind of a scripting tool that kind of pops things up. But to be able to, you know, agency how they're, you know, their words, their tone, their pacing affects customers. I really, really cool. I know that that's the next level of. of what's coming there. But again, it a lot of a lot of and a lot of technical companies that are focusing on this that are that are doing. it. It's not directly into your CCaaS. This one is just a pain for me. I I want to see what skills are the without me having What have we spent today, forecasted cost by channel, by Q, by skill, what is the projected month ends cost?

22:13And again, I can build that in Excel, I can take that the data and pull I can pull it specific customer. All of that that stuff and that's probably more of a BPO thing, but again, to constantly looking at how we're utilizing our agents, you're trying to identify waste better. Hagging anomalies things like that.. But there's still, you know, depending on what a customer wants.. We still have agents that maybe take calls from different customers based on a type of day. Maybe they have analytics on their platform. The other thing I hate is why can't we just use what we I pay for what we use. Thanks for watching! Like again, that's the whole CCaaS, everybody does it, and I think it's kind of garbage. We don't-t have to do it, they want to do it because they make really can explain that well about So the question would be, you know, instead, instead of waiting afterwards, to be able to and that ties into agent assist, to to tie the Q, A, and that together.

24:58I think that that is coming in probably in the next year or two. You know, I love what we're doing with redo. Like to be able to re have agents retake the calls, t they struggled with or they didn and do or had a difficult to re take that call with the AI is something that should be in, in, in, in, And then to be able to build libraries like we are doing with Redo, to keep the chatbot thing that everyone seems to be doing. And again, no differentiation. Pretty boring. I don't know. So anyway, this is my list I'm going to I, I'm sure I'll be adding to it next. I'm for sure it's a pretty chill to a week you guys know the week of Christmas. So I'm going to talk about kind of what is real what is not real. I think in 2026 when it comes to for AI.

26:20And I hate saying that, budget for AI, a I Thank you guys. I apologize for not being, being on as much. I mean, it's been a couple of. months. Like, it's bad. I've been really bad, I

[Gap 26:46–27:16: no reliable transcription]

27:16So anyway, let me know if anybody's listening to this video, please follow me from advice from a call center geek is there. So check, check it out on TikTok. We have probably 300, by all the intelligence available in your, yeah. Yes, so I mean, that's kind of what we, we're saying, right? To have the intelligence of everything that you own, right? To be able the, and a, and again, that's the promise that we were told, right? With AI, we were told, you know, it's of going to be able to look at, you know, from a customer facing, from a customer standpoint, all the CRM data, all and notes, to It could then be able to kind of predict, you know, where this thing should go or why the customer is calling. And I think if it. If it can do that, then there's no reason it can't get into our contact center stack, right?

29:04Everything from workforce management to the CRM to the ACD, to any analytics platforms, any analytics tool, to have all of that data set and then start to make really good decisions on top of that. That's the AI people want, right? Again, maybe we're just not there yet, but I The AI that everybody wants is the chatbot, which I think some people want that. I think that's one spoke of the tool, and it's a pretty important one. There's all the stuff that's being left on the, for no reason, nobody wants to build. I think if somebody actually starts to do that stuff, they're going to be massively successful, especially in the next 3 to 5 years. And again. going to be available in a 2027, 2028. I mean, I don't know, maybe there's a ton of them, or maybe they all just kind of or easier. Instead of just saying, hey, our ROI is the going to be there or we're going lower costs, but how about we actually use AI for other things I mean the ROI is going to be there, but how about we make life easier

[Gap 30:41–30:55: no reliable transcription]

30:55you guys have any more questions. I'd love to talk more about this and I'll see you guys next week. I think next week you'll find very interesting as well. What I think in 2026 is real from what people are saying and all these the vendors go to CCW, what is real, what is not real and what you probably should be investing in And there's a lot of it. I'm not an anti AI guy. I'm for it. But we just got to do it the right way. Appreciate you.