Automation featuring Chafik Abdellaoui

Transcript

Rob Dwyer (00:02.235)
Thanks for joining another episode of Next in Queue. Today I have hurricane survivor Chafik Abdellaoui. Thank you for joining and taking the time out of your day when you could be like, I don't know, cleaning up outside, but here you are with me. I'm glad you're safe. Thank you.

Chafik Abdellaoui (00:24.91)
Thank you, Rob. Thanks for having me. Yeah, I definitely rather be with you than be outside cleaning or doing anything that we've been doing for the past several days. I'm glad we're all safe. Community has been amazing down here. And I think the cleanup is towards the end of it right now. But yeah, I'm glad we made it all safe. And thanks for having me.

Rob Dwyer (00:47.451)
Yeah, absolutely. Just so that people have some context, we're recording this in mid-October. And you did just literally go through a hurricane. And I smile as I call you a hurricane survivor. But it really is a serious deal. And I don't want anyone to think that I'm making light of what just happened. So I really, truly.

hope that everyone in your community is safe and that you guys are working on getting things back to some sense of normalcy because it can be a it can be a really terrible thing to go through. I've seen that kind of aftermath before.

Chafik Abdellaoui (01:28.11)
it

Chafik Abdellaoui (01:32.536)
Yeah, thank you. I appreciate the kind thoughts there. Rob, it's definitely been an experience I wouldn't want to relive and I'm sure a lot of people down here in Tampa wouldn't want to relive that. Hurricane was close, but it was great to see the community come together right after and even before preparing all the homes and after for the cleanup and waiting for people to get power back and people sharing and helping each other was nice to see. So I think we're...

We're starting to get back to normalcy for sure and schools opened back up today, which is nice. So, yeah, thanks for the kind thoughts.

Rob Dwyer (02:07.323)
Yeah, absolutely. We actually had plans to do this a week ago and you obviously needed to reschedule and I was happy to do that. We're going to however talk about really the future of contact centers. Now you are at Autodesk. Tell us a little bit, number one, about Autodesk and then number two, what you do there just to give people some context.

Chafik Abdellaoui (02:38.655)
So, yeah, I work for a company called Autodesk. It's a pretty large technology company that makes software and applications for engineers, architects, builders, people who make movies like for Disney and animated video games. We basically make software that allows people to make anything.

That's the motto of the company. Our flagship product that's been our initial product we started with is AutoCAD. It's a pretty popular product that is used by architects and people in the construction industry and design industry. And then after that, through creating new products, acquiring companies that have cool products, we kept growing over the years. And now we have several products that

allow people to literally make anything. So that would be Autodesk. And myself, I joined four years ago in the Contact Center Technology Department, helping with contact center administration for telephony, CRM, workforce management, quality management tools, and so on.

And then I moved into a role two years ago where I'm leading the workforce management and the quality management functions for our sales teams.

Rob Dwyer (04:06.277)
Yeah, it's a fascinating company. know that anyone who has been involved with any kind of engineering design or architecture design or anything has probably heard of AutoCAD. It's been around for a long, long time and is definitely a big name in that industry. It's the name, if you will.

So...

You and I talked a couple of months ago and you told me then that you really thought the future of contact centers were going to be humans talking and everything else automated. Tell me more about that. what, what did you mean by that?

Chafik Abdellaoui (04:57.486)
So automation has been around forever. So humans have been looking to automate things for a very long, long time. And in the contact center work specifically, we've continuously automated stuff, whether it's data flow from telephony systems to the CRM, whether it's specific tasks of escalations for cases from one team to another team.

Whether it's automated email templates, automated dialers for outbound to help the humans in these contact centers do more faster and better. So automation has always been part of the game. And the latest technological innovation, specifically AI and LLMs that came out that really picked up in 2023 and have been gaining a lot of traction in 2024.

have sped up that process. And now that we have technology that understands human language, large languages, they allow us to automate more things than before. So the idea here and the vision is that within the contact center world, agents will continue having conversations with customers, will continue providing value to the customers that they support, whether it's a sales environment.

or a customer support environment. However, most of the tasks that they have to do today, like this positioning calls in the telephony system, entering information in the CRM, putting in notes and organizing them so the next person who opens that case knows what it's about. All of those things, drafting follow-up emails, sending follow-up.

calendar invites or looping in the right people into the information that needs to be shared. all of those things that the reps have to do either during the call or after a call, will start getting automated one by one. Knowledge will start being pushed automatically to the reps as the machine is listening to the calls and providing reps recommendations on what to say, what questions to ask. And the agents will be empowered by the machine.

Chafik Abdellaoui (07:20.376)
to have the best conversations possible with their customers, which would maximize outcomes, whether it's first contact resolution, customer satisfaction, and so on. And with the reps focusing on that, everything else I see the machine at some point picking up and automating for those reps. So that's what I meant by that.

Rob Dwyer (07:42.489)
Why not, yeah, I love it. Why not just cut out the human and use the machine to do the talking as well? I there are certainly companies that are pursuing that. So do you think that's not going to happen? Do you have an objection to that? Why don't we just automate the whole thing?

Chafik Abdellaoui (08:05.24)
That's definitely something that a lot of companies are pursuing. And I think there's opportunity there to automate the whole thing as long as it's done right. And what I mean by that is humans, us, customers, we've been used to when we call, we get an IVR and it's a small level of annoyance that we have to go through the call flow before we actually connect with someone who's gonna help us. When we go on a website,

Sometimes they hide phone numbers. That's a small level of annoyance that then puts us into a situation where like, okay, I have to search. How am I going to get help here? When we start chatting, it happens to be a bot. And that's a little level of annoyance again, that adds up of what now we're going to chat with the machine before I maybe get help from a human. So those levels of annoyances add up. What I mean by it has to be done right is when you have companies that do it very, very well.

where when you want help, they automatically pop up that chat bot right there to help you. And when you ask for help from that chat bot, it does very well. It gives you the answer that you want. It resolves your issue. And then you move on. That's what needs to happen. And that's where companies need to have first the confidence that they can do it right. And two, the talent, the skills, and the technology

to make sure that it's gonna work the right way. The bots have existed now for a while, whether they were rule-based and now they move AI-based and a bot is only as powerful as the knowledge and the data that it has access to. So for a bot to do a good job, one, the process needs to be straight and working 99 % of the time. And other than the process working, that bot needs to have access fast to accurate data.

in terms of the customer's data and accurate information in terms of the process it needs to follow and how it should answer questions every time a customer needs help with something. We are not there yet, unfortunately. All these demos that you see, all these companies that say we're moving to full automation, full bot will do all this stuff. It's rare.

Chafik Abdellaoui (10:25.748)
when you see it working so well. I've seen it working very well in some companies, but it's rare because of all those challenges and the complexities that come with making sure the bot has everything it needs and is configured the right way to provide outstanding customer support to customers without having to connect them to a rep. Until we get there, I do see the bots still being there, being that frontline that

tries to help customers, but then if it can't help the customer instead of frustrating the customers, route them to a human who will continue helping them. That's the way I view the bot evolution here.

Rob Dwyer (11:05.669)
Yeah.

Rob Dwyer (11:09.809)
One of the things that you just touched on is ensuring that the bot does things correct way. And part of that is access to information. Certainly for larger companies that have more robust knowledge base, they have documented all the things over the years, it does make it a lot easier.

The vast majority of businesses that I come into contact with, they have very little documented when it comes to things like that. so I think that's where the humans are still.

superior because they can still thrive in an environment where things are maybe not documented the same way. Whereas if you're going to automate it, you need to have your docs in a row from a documentation standpoint, and you have to make sure to keep up on it. Like that part is, I think the biggest nightmare for companies is just thinking about how do we maintain our

Our knowledge base? I mean, that's always been a problem, right? How do I maintain my knowledge base? Just period. And make sure that it stays up to date. It's a challenge.

Chafik Abdellaoui (12:26.99)
Yep.

Chafik Abdellaoui (12:34.424)
Yep, definitely. And that's going to continue being a challenge. And even at large companies, it is a challenge too, just because of where, yes, you're right. Large companies have a lot of data, but unfortunately, because they have a lot of data, it doesn't always sit in the same spots. It's not always formatted in the right formats, not always clean. And it's not always up to date. You need the data to remain up to date. And then you need configuration at the bot level to tell the bot, only use data that is up to date. Do not hallucinate.

Do not guess, do not use information from the internet. Only use information that we give you. And the challenge with that is if we're telling an LLM bot only use the information that we give you, then it starts limiting its abilities to be the powerful LLM that it is. But from a business standpoint, we don't want it to be pulling information from everywhere. We want it to say things that...

Rob Dwyer (13:06.309)
Ha ha ha.

Chafik Abdellaoui (13:31.884)
We believe from a customer experience standpoint, from a branding standpoint, from a legal standpoint, from an HR standpoint, from a multiple standpoint, it's not going to do something that's going to cause chaos afterwards, whether it's a brand nightmare on social media, because the bot said something, as you've seen for Air Canada, as you've seen for multiple companies that just went through an LLM bot that's customer facing and then started seeing that.

There's a little bit more configuration, a little bit more control that needs to happen for it to make sure it's going to function properly. So yeah, I think knowledge management will definitely be something that will continue being needed. And then accountability, when companies put these bots and make it harder for customers to reach a human. Don't hold the customer too long if the bot can help them. Great. If the bot cannot help them.

Don't keep asking them, I didn't quite understand that, can you ask me again? Or don't do that stuff. It's like you get one shot, help the customer if you can, move them to a human who's going to help them. Otherwise you're just negatively impacting your brand as you do that to your customers.

Rob Dwyer (14:49.017)
I've just discovered something through this and that is apparently that I should have always just been telling my LLM not to hallucinate and that works apparently. Is that the secret? Hey, by the way, when you do this, do not hallucinate.

Chafik Abdellaoui (14:59.31)
You gotta tell it in many paragraphs. You gotta put so many guardrails. I know.

Rob Dwyer (15:10.117)
We've been trying to solve this problem, Chafik, and now I just learned like the, you just solved it. Like just put that in your prompt and by the way, do not hallucinate. It's genius.

Chafik Abdellaoui (15:18.35)
I

Yep, those guardrails need to be added and sometimes they need to be added in the simplest format and repeated. But as you know, with the challenge, once you start giving so many directions and guidelines to the LLM, you may start losing track of all of the stuff you've given it because it's so much. And then if there's stuff that conflicts with each other, now you're just confusing the LLM.

Rob Dwyer (15:31.494)
you

Chafik Abdellaoui (15:53.88)
So it's definitely a work in progress. And as I mentioned earlier, requires talent, requires people who know what they're talking about and what they're doing, and requires a lot of testing, fine tuning, and continuous improvement.

Rob Dwyer (16:08.453)
Yeah, absolutely. I think there, so one of the things that still concerns me to this day is using LLMs in customer facing situations. And I'll give you an example. I just saw earlier this week is a video. The name of the video was something to the effect of gas lighting.

chat GPT to say 2 plus 2 equals 5. And it is this back and forth between the user and chat GPT who is constantly correcting the bot on what 2 plus 2 equals, but providing incorrect information, literally trying to get it to provide the wrong answer.

And this back and forth ensues, right? And ChatGPT has seen this pattern of 2 plus 2 equals 4 so many times that that is the prediction. But you can, if you work hard enough at it, often get those bots to produce something else, something that you want that maybe is a counterfactual.

And I think that is the biggest risk when you put something in front of customers who may choose whether it's malicious intent or they just want to try things out or whatever the motivation is to try and get a result outside of what the company wants the results to be. So do you think that

Number one, companies are thinking enough about that. But number two, are we going to get to a place where you think, yeah, I would be comfortable having customer facing chat bot answer all kinds of things. And maybe we don't need a lot of human intervention.

Chafik Abdellaoui (18:27.736)
So it depends on the, on the fields. So sales is different than support and support has so many complexities in it. There's some things that are simple support level, some things that are more complex, some things that are pretty transactional, some things that require a lot of judgment calls and making decisions as to how to proceed, which are not so straightforward. I think the bots will eventually continue taking on things.

And they start with the simpler, most transactional things, items, and they still pick up more complex as they go. Just because they're getting smarter and we're getting more possibilities in terms of configuration and things that we can make the bots do. So I definitely think they're to pick up more things and leave in the more and more complex stuff to the humans. So contact centers in terms of their sizes, potentially.

It could decrease. It all depends also on the demand outside of that company's products and all of that. So, but overall contact centers that implement these technologies, they can get to a point where they can get one agent to do the work of 10 agents and also the front customer facing bots to take on most of the work that is

transactional that the bot is effectively able to do. So that's a natural trend in contact centers. Now with that automation, people have been talking about this for before LLM, that automation will replace agents, automation will decrease contact centers. Automation is helping with that stuff, but the other side of the coin is the market is big enough. There's a lot of companies.

There's a lot of services, a lot of products, and there's a lot of consumers out there. So as long as demand is growing globally and people are there and we needed to help contact centers in terms of employment, in terms of the number of people working there could potentially still grow even with the automation that is happening. So it's not just one side of the coin. It's a giant market. It's a giant industry globally that has.

Chafik Abdellaoui (20:55.904)
millions of people working in it with the agents being the front line of those associates in the industry. I hope this answers your question, Rob.

Rob Dwyer (21:07.949)
Yeah, absolutely. What I think is fascinating and I think is not talked enough about, talked about enough is that there is a good chance that all of this automation doesn't reduce the number of agents that we need worldwide. We certainly have been automating things in contact centers for decades and yet

business continues to grow and grow and grow. And I do think that some people look at AI and automatically say, well, that's it. That's the end of contact centers. By the way, the same thing they said about IVRs and chatbots. And yet here we are. There have been some particular situations with some particular companies where they've been able to reduce head count.

in their contact or customer support functions. But I don't know that that is going to be the trend from a global standpoint with all of these companies. And part of that is simply boils down to the fact that you have a lot of small companies that are still going to have the same struggles that I talked about earlier where you need people because the things just aren't very well documented or things are so fluid because it's a young company, a startup company that's growing.

Chafik Abdellaoui (22:27.746)
Yep.

Rob Dwyer (22:36.355)
that things are changing all the time. And it's just easier to manage that with people than it is with an LLM that has also potentially can't forget. Because that is another challenge that LLMs have about them is once you get data in, sometimes it's really hard to it to forget that data and not utilize it anymore.

Chafik Abdellaoui (23:00.443)
Yeah.

Chafik Abdellaoui (23:05.102)
Great, great points. And I'll just add to that, a very basic, basic, basic thing is the human population in the world continues to grow. So just with that, there's gonna be more people that need more services, more products, and these contact centers, just like you said, they keep growing despite automation happening for decades and decades and decades.

Rob Dwyer (23:05.328)
Let's.

Rob Dwyer (23:29.583)
Yeah, if we have as humans have figured anything out, it's how to continually increase our numbers across the globe. We're very, very good at that.

Chafik Abdellaoui (23:37.91)
Yep.

Rob Dwyer (23:43.513)
I want to talk a little bit about AI versus automation. And is there still a place for old school automation? So I'll just start there. Can you break down the difference between the two? And then what are your thoughts? Is automation going away or

Chafik Abdellaoui (24:08.866)
No, I think automation is a basic. It's the process that will continue. Automation is not going away. It's something that existed for a long time and will continue existing for a very long time in the future. I think AI is an advancement that allows us to automate in easier and faster ways. And the automation that we were used to in the past, say solid 20, 30 years has been rule-based. Rule where...

Give it a decision tree, you give it information and it based on multiple scenarios and conditions. If this equals X, this equals Y and this equals A then do this. That's been the rule based automation that we've been implementing in systems, whether it's an IVR, whether it's a workflow between a CRM and a telephony system, whether it's an escalation, anything in the contact center industry. I think the AI piece, the artificial intelligence,

And specifically within the AI, what impacts the contact center the most is the LLM, the large language model, not the full AI with knowledge of everything in the internet, but the LLM that now makes the machine able to clearly understand what humans are talking about. It's not just capturing keywords. It's not just transcribing and understanding.

putting in text what we're saying, it's actually understanding what we're talking about, where it can tell you from the conversation that Rob and Chafik had, this is what they discussed, these are the cool insights that got them excited to discuss more, these are the questions that were raised, it's smart enough to do that. And this is a major advancement because what that means is now it's smart enough to understand conversations between agents and customers.

It's smart enough to understand information in the CRM, notes, emails. It's smart enough to understand what's happening and what next steps could be coming after what's happening. Because of that, it's making automation easier, faster, and more effective for the companies in the contact center industry that want to leverage the LLM. But automation will always be there. That's my...

Chafik Abdellaoui (26:28.984)
point of view at least on this.

Rob Dwyer (26:32.593)
I also wonder, speaking of LLMs, and this may be something that the general public is not aware of, but when it comes to a large language model, a lot of the work is being done in specific languages. Obviously, English is a huge language, probably dominates the market, but there are quite a few other languages where LLM work is being done, these models are being created, but there are other languages.

where there's not much being done. They may have a smaller population base. Those languages and the people who speak them

they still need help, right? There is still an opportunity for contact centers to thrive in the human-based aspect and even being a little bit old school, simply because it's going to take a while, if ever, right? Someone has to decide that it's worth the investment to put that into specific languages. And the nuances of certain languages can make

getting an LLM to perform well, challenging. There are lots of different nuances within specific languages that make understanding even for a person different than some other language. My go-to example, only because I happen to know a little bit about that language, is German, right? The verbs are always at the end. That's weird. But that's

That's Germans for you, right? You don't know what they're doing until they get done talking. And then you're like, OK, now I know what you're doing. So there is a challenge globally when it comes to the efficacy of LLMs because of language barriers, which is itself a little bit of an oxymoron here. We're talking about LLMs, but they do have language barriers, right?

Chafik Abdellaoui (28:39.854)
Yeah. So we've been, we've definitely been using LLMs for multiple languages at Autodesk. And you're absolutely right. It does better in some languages compared to others. Now I would say technology, AI, LLMs, they follow capitalism and wherever the market is big, wherever there's a lot of conversations, wherever there's a lot of content online, the better they operate. So.

in the big markets where we do sell a lot and where the language is pretty big, I've seen LLMs do very well. Even for German, I've seen it perform pretty well in terms of what we're asking it to capture from conversations and what information we're looking to automate workflows for. So for those dialects, like I'm originally from Morocco,

and the Arab world is giant, goes from Morocco all the way in in northwest Africa, all the way to the Middle East. So almost one and a half times the size of the United States in terms of all the countries and many, many dialects across the Arab world. So I definitely do not see LLMs do well with those Arabic dialects. However,

I see it do well with the classical Arabic where that's the Arabic spoken on Al Jazeera or like new, you read in newspapers or whatever, because there's enough content online in that language for it to perform well. So for specific dialects, I definitely think that there's ways to go just because AI doesn't have enough content to work with and learn from. But for the big markets, the...

EU, Japan, Korea, United States, Canada, Brazil, Mexico, the big, big markets. I think AI is pretty powerful today and will only continue getting better to allow companies to automate more processes and generate higher revenues from these big markets. And in capitalism, usually companies focus on these big markets. They do every now and then go into the smaller markets and Africa and Southeast Asia and so on. But

Rob Dwyer (30:54.075)
Yeah.

Chafik Abdellaoui (31:04.206)
they focus on China, they focus on India, they focus on the big giant markets for sure. And in Chinese, it's AI is very powerful.

Rob Dwyer (31:12.602)
He just sent it.

Rob Dwyer (31:16.463)
Yeah, you just sent half my American audience to Google Maps because we are famously awful at geography. go, go find Morocco on a map. You'll understand what what Chafik was talking about. I want to dig into a little bit about your, current expertise and just ask you how AI is impacting

workforce management. So for the, for the uninitiated, which actually the people that listen to the show, I think are probably really good at geography because a lot of them have been involved in contact centers and or BPOs. And they probably know what WFM is for those who don't. These are the people who basically handle schedules and workforce demands. They

determine how many people need to be available at any given time to make sure that you're not waiting 25 minutes on hold to talk to someone. But has AI impacted that function? And if so, how?

Chafik Abdellaoui (32:33.198)
haven't seen it impacted too much now. I've been participating, like just attending demos and podcasts, especially led by the WFM vendors, whether it's Genesis or Nice or Calabrio, Verin, or all those guys. And they do show some of the AI stuff that they're implementing within the WFM tools for scheduling, forecasting, and so on. It's not as much as what you see in quality.

or in the CRM part of the contact center. I definitely see opportunity there. see in the future, the intraday management piece of WFM being more more automated. We do not need people watching queues all day long and looking at agent adherence all day long and making intraday changes within the next hour and things like that. To me, once we've made our plans,

once the day has started and we have good discipline within the contact center, the little changes here and there to me can be automated. And I do see AI if these vendors implemented properly within their tools to start helping with intraday functions. So that's where I see the opportunity, but I haven't seen it in action yet. So my...

start would be that intraday management would be the first to go and be automated, or at least a big chunk of it to be automated. The rest, the more long-term planning, relationship building, relationship management with the managers, the leads, the agents, the general WFM processes for reviewing last week, reviewing yesterday, reviewing next week, reviewing long-term, next month and so on. I don't see that stuff being automated that much.

just because of all the variables that you're aware of coming into play when it comes to WFM planning and the general WFM function.

Rob Dwyer (34:40.187)
Well, and as we've already discussed, LLMs are famously bad at math. You can get chat GPT to tell you 2 plus 2 equals 5, which is not going to be good for your long-term forecast if it's using 2 plus 2 equals 5. That's probably going to mess some things up. Am I right?

Chafik Abdellaoui (35:00.59)
Yep, mean, that's definitely the the caricatural view of AI. But yeah, I haven't seen it do good in terms yet, in terms of things like long term planning, where you have to put in a whole bunch of assumptions and variations and stuff where you have to have discussions with people, you got to bring in finance, you got to bring in HR leaders, operations, various stakeholders just to come up with a solid plan and then

work your way through the plan and see how the company is doing against that plan. It's not stuff that AI is able to help with now. And even if it was, there's going to be the, how do you say that? People not having trust and people questioning things. People are already questioning things now when they're working with people. as you know, so imagine if it's the machine telling them things regarding long-term plans and...

You need 10 extra people or you're over staffed right now or your attrition is not what it's supposed to be. I don't see the machine being able to handle a lot of the WFM function just because of how complex it is and how it requires that trust and collaboration and exchange between people in order to make business decisions.

Rob Dwyer (36:21.369)
Yeah, it almost strikes me as you're talking about that two things. One, particularly LLMs, they're really good at recognizing patterns. And so when you're looking at past performance, almost like the stock market, the saying is past performance is not a predictor of future performance.

And that's kind of why WFM is a little bit of art and a little bit of science because, you are using the past to help you inform the future, but you start talking about all these different variables. And that's where the art comes in and understanding like, okay, well, you know, my expectations, for instance, for the weather, we started this talking about the hurricane weather.

can have a huge impact on call demand depending on the industry that you're in. So you and I offline before we were talking, you made the decision to stay put in part because of traffic, inability to find a hotel anywhere close. And would you want to talk about volume going through the roof? Hotels in the Southeast,

where long-term stays in the Southeast when hurricanes are active, their volume just goes through the roof, right? And that is not a constant like a holiday is. I know that Black Friday, yes, volume's gonna go up, and the week after volume's gonna go up. The week after Christmas volume's gonna go up, right? These things I can potentially predict really well.

But when you start throwing in variables like weather, it's really challenging. But the other thing that strikes me as you were talking through someone coming to you and saying, hey, you've got too many people, blah, blah. We like to hold people accountable. As leaders of business, we want to have someone that we can

Rob Dwyer (38:42.577)
hold accountable. How do you hold AI accountable when it messes up? Like, what are we going to do then? Is that something that you think that will drive a lot of retention of humans in roles that maybe AI could do an okay job of, but we just don't really have anyone to hold accountable?

Chafik Abdellaoui (38:50.21)
Yep.

Chafik Abdellaoui (39:06.594)
Yeah, the machine would never be held accountable when you have a machine that makes shampoo in a factory or a machine that makes soap or whatever. If the machine breaks, you don't go and start yelling at the machine, why did you stop production? You go find the people who know the machine. We're like, how can we fix it fast? So same thing with AI, yeah, just a machine.

Rob Dwyer (39:25.157)
You lazy machine! Get back to work!

Chafik Abdellaoui (39:34.488)
Configure it we buy it we use it we give it information if it messes up. It's the people who are in charge of it who? Who are held accountable so I? Don't think we'll get to a point anytime soon where AI is in charge of major decisions that Really get us into the talks of accountability like go hire 10 people because you need 10 extra people and then

You go hire those people and they're just rolling their thumbs because there's no calls coming through or whatever because AI told you. I don't think we're to get to a point where we're going to trust the machine with big decisions like that. As I mentioned earlier, we still need to have collaboration discussions, people throwing in their point of views, people sharing data, people providing their opinion based on experience, based on their gut feel. There's things that the machine is not going to get to anytime soon, but.

For everything transactional, everything fast base, everything where it's predictable decision-making. I think we will get to a point where AI will be helping with all of that and automating processes because we don't need to sit in a meeting or exchange on why we need to remove the big training from this afternoon. If AI sees that the call volume is spiking because there's a hurricane in the Southeast and we have seven reps who are scheduled for team training.

We can have the machine automatically remove that training, send a notification to everyone and be like, yo, there's a thousand more calls per minute and we just can't afford to have people on this meeting. So we'll reschedule it to another time when volume is more manageable. We don't need humans to have a meeting to make that decision. That's, that's what I'm saying. And that we will get to, but anything bigger than that, where we're talking about staffing, where we're talking about things that impact our customers, like.

how we're going to deal with certain inquiries or whatever. AI may help us with the decision making a little bit, but it's not going to make the decisions anytime in the near future.

Rob Dwyer (41:44.207)
If it does, I want to be the fly on the wall when AI is pulled into the HR office and is terminated for cause after multiple failures and being on a warning. That's what I'm looking forward to. It seems like there's a meme there somewhere. Chafik, I look forward to seeing what

Chafik Abdellaoui (42:04.748)
BLEH

Rob Dwyer (42:11.833)
AI is able to help us do. I know that you are staying on top of it and I can't wait to talk to you again. We're going to have to. We're going to have to maybe have. I have this feeling we're going to have to do some kind of roundtable event with all stars like yourself. I haven't figured it out yet. I'm just throwing things out there so I'm going to invite you when I put that together.

Thank you so much for joining the show today. If anyone wants to tell you that you're full of it, or maybe they want to come clean up your yard and help you with that, what's the best way for them to get in touch with you?

Chafik Abdellaoui (42:57.518)
Thanks for having me and thank you so much for any future invites in advance. I would love to participate in those. If anybody wants to reach out to me, they're welcome to connect with me on LinkedIn. You're also able to share my contact information with anybody who reaches out to you if they want to connect with me. But yeah, I'm always open and looking forward to connecting with people in the industry and people who are passionate about automating things, passionate about

contact centers, WFM, quality, knowledge management. It's just a field that I'm very passionate about myself. And if there's people who share the same passion in the contact center world, I'm always looking to connect with them.

Rob Dwyer (43:42.012)
Absolutely, well, you know the drill folks head on down to the show notes. You'll find a link to Chafik's LinkedIn profile. Give them a connection request. Say hello, tell them I sent you. Chafik, thank you for joining Next In Queue.

Chafik Abdellaoui (43:59.704)
Thank you for having me Rob, looking forward to chatting again.