Menu
February 4, 2019 | By

AI IRL Podcast Episode 5: Why chatbots are here to stay


Subscribe via iTunes, Spotify and more.

Chatbots are everywhere these days.

It seems like you can’t log into any sort of SaaS website without seeing that little popup window in the bottom right-hand corner of your screen saying something to the effect of:

“How can I help?”

But are those chatbots actually helpful? Are they actually doing what the customers are wanting? And is there a limit to what a bot can help the customer accomplish vs what an actual human can do?

On a recent episode of AI:IRL, we sat down with Mike McGarvey, Account CTO & Sr. Director at Atos and talked about just that.

What is happening in the AI space in terms of chatbots? Where is the chatbot technology going? And how might your company better set up a bot to help customers get the help that they need, all while they’re still doing other things?

Allowing for MultiTasking

“We’re finally getting to the point where chatbots are becoming able to break apart intent and really analyze a customer’s desires,” Mike says.

Once you’ve deployed a chat capability, you can begin to run change management and really get people to adopt use of the bot.

And once you’ve got a good handle on what people are coming to the channel for, what particular problems they’re asking the chatbot to solve, then you can begin to customize your bot and layer more and more intelligence on top of it.

Let’s face it. Chat has been a feature of the internet for the last 15+ years.

Unfortunately, for most companies, chat has been an afterthought, often even in the single digit adoption rates, which is a shame.

More and more end users, from all walks of life, but particularly in a corporate setting, are utilizing chatbot technology to multitask.

Think about it this way: If you’re on a conference call at work, but you know you need to get in touch with your cable company with a question about your bill, how much more likely are you to utilize a chatbot, knowing that you can continue on your call, while still at the same time, be negotiating your cable bill?

Now imagine that you know that if the bot can’t help you, it can escalate your problem to a live representative who can chat with you and help you fix your problem.

It’s a no-brainer.

What’s Next?

AI is very good at knowledge retrieval. At saying, “I see you input this information, allow me to return this information to you.”

But according to Mike, over the next 6 to 18 months, the future of chatbot technology is in DOING things.

“It’s not just about the simple things, which is sharing knowledge, it’s about the complex things where we actually DO work.” Michael McGarvey

But what does this look like? Let’s look at it through the example of planning a trip.

Knowledge Management

  • “Where is my flight taking off from?”

Assistance

  • “I need to check in 24 hours in advance. Please help me do that.”

Agency

  • “I need to plan a trip. Here’s my budget, timeframe, and location. Help me book a trip.”

Most companies are likely the knowledge management spectrum, maybe the assistance. But the AI technology isn’t quite to agency yet.

But it will be.

Personalize Intent

As we move forward in the chatbot space personalizing intent is going to become of utmost importance.

Even something as simple as analyzing a customer’s bill over the prior months and looking for increases. If a customer’s bill has increased 10% or more over the last month, there’s a good chance that that’s the reason they’re coming to your channel.

So starting the conversation out with, “I see your bill increased recently. Would you like help understanding bill details?” could be a massive shift in the chatbot space.

If your company isn’t currently utilizing chatbots for customer support, it’s likely that you’re either currently planning bot integration, or that you will be soon.

 “I think over the next few years, our interaction with chatbots is going to become second nature.” Michael McGarvey

Every business wants to adapt, save time, and allow for quicker interaction with customers. Chatbots do just that, and with the right data and intelligence behind them, can save you and your customers an untold amount of time, energy, and headaches.

So what are you waiting for?

This AI discussion with Mike McGarvey was taken from our podcast.  If you want more AI: In Real Life, check out our podcast on iTunes.

View transcript »


Ryan Lester
00:05 – 00:12
Welcome, this is Ryan Lester your host of AI in real life joining me today in our weekly podcast is Mike McGarvey.
Ryan Lester
00:12 – 00:15
Mike’s works with Otto’s Mike.
Ryan Lester
00:15 – 00:15
Welcome.
Ryan Lester
00:15 – 00:16
Thanks for joining.
Ryan Lester
00:16 – 00:17
Thanks for having me.
Michael McGarvey
00:17 – 00:20
Looking forward to our conversation today.
Ryan Lester
00:20 – 00:24
I know you and I have worked across a number of different projects with a variety of different companies.
Ryan Lester
00:24 – 00:34
What excites me about your role is really get your hands dirty in a lot of different projects that can you start out by talking a little bit about your current job and some of the work you do on a day-to-day basis.
Ryan Lester
00:35 – 00:35
Yeah.
Michael McGarvey
00:35 – 00:36
Sure.
Michael McGarvey
00:36 – 00:55
So my role as an account CTO, I work with clients basically meeting with them to figure out how we can solve business challenges with Solutions and my role within auto switch is a global many services companies really incubate new technologies, which we can roll out to other accounts.
Michael McGarvey
00:55 – 01:03
So we kind of get a common ground common increase in the number of applications and that was awful.
Michael McGarvey
01:03 – 01:04
I’m sorry, that’s fine.
Michael McGarvey
01:04 – 01:05
Yeah.
Michael McGarvey
01:05 – 01:14
Restart will just hit take a pause will cut and then you can restart again just with your you know, I’ll ask you the question again, then you can go again awesome.
Michael McGarvey
01:14 – 01:21
So Mike welcome watch them a little about your current role at Otto’s and what keeps you busy day today.
Ryan Lester
01:21 – 01:22
Oh, thanks.
Michael McGarvey
01:22 – 01:24
So my account.
Michael McGarvey
01:24 – 01:42
My role is the count CTO and I work with customers to kind of solve very interesting business cases and develop new use cases based on technology Integrations new development projects, but we really where I focus on is digital workplace and coming up with new solutions for our customers and incubating new products for autos.
Michael McGarvey
01:42 – 01:50
Yeah, I think what’s really interesting for me is you kind of a to your interesting one foot in two different camps of one the customer Camp.
Ryan Lester
01:51 – 01:56
So what’s the business Challenger trying to address where they and their kind of technology or business process journey.
Ryan Lester
01:56 – 01:59
And then also you’re doing that even within Auto so there’s those best practices.
Ryan Lester
02:00 – 02:03
You guys are learning across all of your different client base.
Ryan Lester
02:03 – 02:12
So I think it’s interesting for you to kind of be a practitioner not only within your own company, but also then drive that practition engaging or any of those those practices out to your clients.
Ryan Lester
02:12 – 02:20
So, you know every time you and I chat I always really kind of enjoyed that perspective of you can kind of incubate it and then drive it out into real production.
Michael McGarvey
02:20 – 02:29
Yeah, we find that really if you’re not embedded with your clients trying to find for new Solutions and you’re looking in the wrong place.
Michael McGarvey
02:29 – 02:30
Yeah, absolutely.
Ryan Lester
02:30 – 02:42
So I know you and I’ve talked about AI number of times and it’s obviously a big Hot Topic lots of ways companies are looking at AI, you know in the conversations you’re having I want to start off and just
Ryan Lester
02:42 – 02:44
Do you see what’s the current Trend?
Michael McGarvey
02:44 – 03:08
Where do you see current levels of adoption kind of what’s this first approach or how a company’s first approaching AI or at least where they and their AI Journey with the client you’re working with well specifically around you know chat AI it seems to be we’re finally getting to the point where you know, the chat Bots can understand and and break apart an intent and really kind of analyze it and match it towards towards knowledge.
Michael McGarvey
03:08 – 03:24
And so really where I think that most of our clients are looking at right now is how do we put a chatbot in front of our clients that are part of our end users that can actually derive an intent find the right knowledge and present it on a very effective way.
Michael McGarvey
03:24 – 03:28
So we’re finding that that is is kind of where the point of maturity level is.
Michael McGarvey
03:28 – 03:42
We’re certainly looking to to grow that but you know from a from a from a best practice perspective sort of where we see this as is you deploy a chat capability you can you run the operational
Michael McGarvey
03:42 – 03:45
Change management get people to adopt it.
Michael McGarvey
03:45 – 03:56
And then once you kind of a good handle on what people are coming to coming to the chat channel for then you can customize your chat bot and really kind of layer that on top of it.
Michael McGarvey
03:56 – 04:01
Yeah, and it’s interesting the way you say that of because there’s actually there’s this interesting thing.
Ryan Lester
04:01 – 04:06
I we see when we talk to companies of there are some who are still like in the early days of even chat.
Ryan Lester
04:06 – 04:16
I mean Chad has been a you know, Technology’s been out there for 10-15 years, but they’re kind of either saying hey we haven’t looked at this in Channel because we haven’t really seen the value.
Ryan Lester
04:16 – 04:22
I’m so they’re kind of looking at chat for the first time or they have chat and now they’re saying okay.
Ryan Lester
04:22 – 04:23
We want to layer a eye on top of it.
Ryan Lester
04:23 – 04:27
And so I think it’s it’s interesting when you talk about best practices.
Ryan Lester
04:28 – 04:31
It’s like they’re still this human element of it’s not like it’s just chat bot.
Ryan Lester
04:31 – 04:38
It’s chat with a human but then layering and on top of that to drive operational efficiency better results, etc.
Ryan Lester
04:38 – 04:39
Etc.
Ryan Lester
04:39 – 04:42
So with most companies working with do they already have
Ryan Lester
04:42 – 04:49
Are they really some of them are looking at both from a we’re going to human chat and we’re going to add AI power chat.
Michael McGarvey
04:49 – 04:55
Yeah, it certainly makes right but a lot of a lot of our clients have chat and sometimes it’s been an afterthought.
Michael McGarvey
04:55 – 05:20
Sometimes it’s been in the single-digit adoption rates, which is a shame because we find that more and more of our end users especially office knowledge workers like that sort of they can have a conversation they can keep doing additional things and they’re not really impacting their day so they can be on a conference call and they can shout the service desk to get something resolved obviously kind of splitting their focus and getting more than one thing at a time.
Michael McGarvey
05:20 – 05:34
But ultimately, you know for the adoption to work well for a chatbot, for example, they need to know that they can go to one channel and if it’s front end with a chatbot the jackpot for whatever reason does have the right knowledge or can’t can answer the right question.
Michael McGarvey
05:35 – 05:39
They can be seamlessly escalated to to and an operator.
Michael McGarvey
05:39 – 05:42
So the operator can come in and kind of have the chat.
Michael McGarvey
05:42 – 05:54
And see what they’ve already talked about avoid having to repeat conversation, but really get to the root cause of the problem quickly and ultimately the end users know I can go to this one channel and maybe have a conversation in the chat bar for a couple minutes.
Michael McGarvey
05:54 – 05:58
But if they can’t help me I’ll go to an agent and ultimately get resolution to looking for.
Michael McGarvey
05:58 – 06:06
Yeah, and I think another thing that you and I have talked about is the fact that this would become more pervasive across channels.
Ryan Lester
06:06 – 06:07
So people are putting into an app.
Ryan Lester
06:07 – 06:09
They’re making it multilingual.
Ryan Lester
06:09 – 06:11
So there’s kind of these opportunities.
Ryan Lester
06:11 – 06:16
I think one of things is driving the growth of this as a channel is that to your point.
Ryan Lester
06:16 – 06:23
It can be multi-threaded you kind of can engage when and where you want and that’s also being supported by the fact that you can do it on your mobile device.
Michael McGarvey
06:23 – 06:42
So when you’re walking between meetings or you’re using your laptop for our conference call or to do some productivity work, you can also be doing this kind of side work on your mobile device or in some other channel to once again kind of increase your productivity by doing a couple things at once now, absolutely, you know, we find a couple
Michael McGarvey
06:42 – 06:47
Can use cases that we have a lot of our clients who have mobile first workforces.
Michael McGarvey
06:47 – 06:54
These are you know reps and sales people out in the field and they need help and sometimes they don’t have the ability to just break pick up the phone.
Michael McGarvey
06:54 – 07:08
They just need to get on there, you know have a quick chat get something resolved and the me on the front end that with hey, how do I do this and getting real quick answer before they go and see a doctor or before they go and see A New Perspective client is helpful.
Michael McGarvey
07:08 – 07:21
So we’ve actually taken that and embedded into kind of a product we have which we roll out to our customers that allow them to look up the status of their ticket.
Michael McGarvey
07:21 – 07:27
They can schedule phone call, but primarily they can shout service ask on their mobile phone.
Michael McGarvey
07:27 – 07:29
Yeah, I mean so much of it.
Ryan Lester
07:29 – 08:04
Once again is just enabling, you know, the way your employees need to work or your users need to work and much this like a lot of things that gets me excited about this space is just not forcing someone into one workflow of like you got to make a phone call, but rather the workflow that works best for the way they work and I think you know AI layers on top of what kind of chat already brought or messaging already brought to make it more real time make it 24/7, you know handle a lot of customers at once versus being constrained by how many agents you have online.
Ryan Lester
08:04 – 08:14
So I think there’s a lot of kind of amplification of some of the goodness that already came with chat and messaging with AI and that’s a great example of it.
Ryan Lester
08:14 – 08:27
So we kind of talked about where we are today with with a I guess as we’re looking at what are some of the additional capability Beyond just the intent recognition and Knowledge Management knowledge retrieval, I guess where our people kind of taking
Ryan Lester
08:27 – 08:31
Next step in that AI journey and what some new things that are emerging?
Ryan Lester
08:32 – 08:32
Yeah.
Michael McGarvey
08:32 – 08:44
So I think knowledge retrieval is kind of the the level where most AI products chatbot products are where they can they can retrieve knowledge and they can present it back pretty well for the for the end users.
Michael McGarvey
08:44 – 08:46
But really in the next 6 12 18 months.
Michael McGarvey
08:46 – 08:53
We’re seeing this emerging field of actually doing things in an automated way with our with our end users.
Michael McGarvey
08:53 – 09:05
So whether it’s interfacing with a procurement system, whether it’s interfacing with our PA products to be able to do things you don’t want to one example, we use is trying to if you come into the chat bot you’re already authenticated.
Michael McGarvey
09:05 – 09:06
We already know who you are.
Michael McGarvey
09:07 – 09:08
We already know things about you.
Michael McGarvey
09:08 – 09:27
Can we use that to qualify and customize intense if you say Hey, listen, I need to unlock my SAP account, you know, then we can qualify do a few things and then run a robot on the back end and clear some fields or or Grant simple permission things that we can do in an automated way, but it
Michael McGarvey
09:27 – 09:34
Comes down to driving in the 10th qualifying what the questions we need to do to execute and then running that execution.
Michael McGarvey
09:34 – 09:38
So it’s not just about the simple things which is sharing knowledge.
Michael McGarvey
09:38 – 09:40
It’s about the more complex things were actually do work.
Michael McGarvey
09:41 – 09:43
Yeah, there’s a construct.
Ryan Lester
09:43 – 09:52
I really like actually got it from a smart person that works right down the street here at Fidelity of talking about this evolution of AI from the first phase is Knowledge Management.
Ryan Lester
09:52 – 09:57
So once again to your point knowledge retrieval the second phase of this isn’t really know about assistance.
Ryan Lester
09:58 – 10:07
So help me complete a task and then the third phase will be more like true agency rather than just completing a task like do complex processes on my behalf.
Ryan Lester
10:07 – 10:17
And the example I was like to think about is, you know, the the Knowledge Management is something like where’s my flight taking off from the assistance is you know, I need to check in the next 24 hours.
Ryan Lester
10:17 – 10:26
Please go help me check in and complete that those few steps involve checking in and an agency is I want to book a trip and here’s my here’s my budget.
Ryan Lester
10:26 – 10:31
Here’s my you know, the timeframe go do all that work for me and you were certainly not at the point yet.
Ryan Lester
10:31 – 10:34
Where are you guys going to complete a whole bunch of complex tasks?
Ryan Lester
10:34 – 10:39
But I think to your point we’re starting to see a shift towards more of this assistance model.
Ryan Lester
10:39 – 10:41
Where what can we plug in?
Ryan Lester
10:41 – 10:44
Intent engine two solutions.
Ryan Lester
10:44 – 10:48
We ask those follow-on question to really understand your state start to understand what the users trying to accomplish.
Ryan Lester
10:49 – 10:58
How can we then help to automate those tasks or whatever’s needed to accomplish that so I think that’s an important point for folks as they’re thinking about.
Ryan Lester
10:58 – 10:58
Okay.
Ryan Lester
10:58 – 11:03
I’ve you know started to play a chatbot or start to play some aii around intent.
Ryan Lester
11:03 – 11:15
What’s kind of the next things I should be thinking about and so a couple things you talked about were qualification plug into other systems may be some sort of authentication if you want to talk more about that.
Michael McGarvey
11:15 – 11:41
What are you seeing as have some core components people need to make this more of an assistance model what I think there is you need to personalize the internet the intent and and that’s sort of where we’re spending some of our focus and our d on now is how do we know and anticipate intent when users come in for example, Ryan, let’s say you’re a new employee and you say well, thank you for joining the company.
Michael McGarvey
11:41 – 11:44
I see the your new employee.
Michael McGarvey
11:44 – 11:44
Can I help you?
Michael McGarvey
11:44 – 11:48
Reset your password or get your PC set-up or can I be something else?
Michael McGarvey
11:48 – 11:49
Right.
Michael McGarvey
11:49 – 12:16
Can you can you you know, give kind of a nice qualified greeting and anticipate the intent and by doing that and pulling in information from various sources, whether they be which application you have access to whether they be the status of are you a new employee what business unit your part of and sort of customize the interaction so that it is more personalized and starts to become Superior to even chat right?
Michael McGarvey
12:16 – 12:19
Because you’re getting opted up and they’re actually able to guess what it is.
Michael McGarvey
12:19 – 12:38
So for example, if your account just got locked out and you have a con if you start trading with the chat Bots and the things as I see your cock blocked out, can I help you and change unlock that if you count just got locked out and you you reach out to a support Channel probably calling about to get it take to get your account as reset upright.
Michael McGarvey
12:38 – 12:41
So trying to end to anticipate that intent use
Michael McGarvey
12:41 – 12:59
Back-end systems to kind of interface with these AI chatbots is really where we’re spending a lot of our focus on recently and I really like that too because to me, you know, the power of AI is it’s like probabilistic models or what’s our what’s the confidence level?
Ryan Lester
12:59 – 13:26
We feel like is based on, you know, a phrase put in or based on the data we have we have a certain sense of confidence of this is what their intent is what their objective is and then the AI then tries to engage in that use case for that dialogue and I think what’s interesting about this is that because you can feed all this data into the algorithms into the AI all these different data sources kind of all at once, you know, a human can only kind of work across so many workflows at once.
Ryan Lester
13:26 – 13:32
There’s this really unique opportunity to get much more to your point predictive or proactive.
Ryan Lester
13:32 – 13:34
There’s an interesting use case.
Ryan Lester
13:34 – 13:41
We were working on with a service provider where you know, oftentimes people will engage when they have questions about their bill.
Ryan Lester
13:41 – 14:04
And they said hey if there’s a certain percentage difference in the bill for month over month and you can calibrate this but it’s a hey if the village Changed by 10 percent month-over-month when a user comes to the site one of the first things were going to ask them about is do you have questions about your bill and then they built a workflow around highlighting why the bill was different and you could ask questions of the bot to say.
Ryan Lester
14:04 – 14:13
Oh, you know, like you can put numbers in a why am I paying $10 more in the bot would understand when you when you gave that number it would look within the bill and say oh here’s why that was different.
Ryan Lester
14:13 – 14:21
And so it’s reducing that kind of low value conversation for an agent asking questions about a bill but doing it in a kind of very proactive way.
Ryan Lester
14:21 – 14:25
We are not asking someone about their bill if they if it’s the same bill as last month.
Ryan Lester
14:25 – 14:32
So I really like this example of you know, the more of that type of data you can feed the bot the better.
Ryan Lester
14:32 – 14:41
I can be it to your point getting ahead of what the customers needs are and then really kind of delaying that customer to answer their question kind of before they know it or
Ryan Lester
14:41 – 14:46
It you know in a very few numbers of number of exchanges back and forth.
Ryan Lester
14:46 – 14:48
So I really like that as an example.
Michael McGarvey
14:48 – 15:02
No, I agree and I think ultimately on a maturity LE model that’s sort of where it’s going to go is that you know, we have virtual assistants and we have chat Bots and I think they’re going to become to indistinguishable in the not too distant future and they’re going to know pretty much anything you anything any data that’s out.
Michael McGarvey
15:02 – 15:06
There will be able to pull it in understand what your world looks like.
Michael McGarvey
15:06 – 15:27
Whether it’s your calendar, whether it’s the status of your access accounts, whether it’s your emails in your inbox and be able to have a natural language conversation with with this virtual assistants last chat Bots and get results quickly and and it’s going to become so second nature to us just in a few years that the adoption will be super high.
Michael McGarvey
15:28 – 15:31
Yeah, I mean I tend to agree with you.
Ryan Lester
15:31 – 15:33
I think it’s an interesting thing of it all.
Ryan Lester
15:33 – 15:39
So I think will help drive that adoption is also the fact that it will become more useful because it will be more proactive.
Ryan Lester
15:39 – 15:41
It’ll be more predictive.
Ryan Lester
15:41 – 15:45
So rather than it kind of waiting for you to come along and engage with it asking a knowledgebase question.
Ryan Lester
15:46 – 15:50
It’ll more say, you know, I see you’re doing this or this is different.
Ryan Lester
15:50 – 15:51
How can I help you?
Ryan Lester
15:51 – 15:52
Or here’s what I think I can help you with.
Ryan Lester
15:53 – 16:01
So I definitely agree that that will drive the adoption and the usage because it will be more value add absolutely.
Michael McGarvey
16:01 – 16:05
It’s also trying to find out how we can expose this as many channels as possible.
Michael McGarvey
16:05 – 16:20
You know it the chat channels obviously a very active one, you know using things like Alexa for business and be able to engage it casually in the work environment putting it into the cloud-based contact center so that when you call the service has given older channels people still call.
Michael McGarvey
16:20 – 16:28
Can we somehow integrate, you know one chatbot whether it be voice chat bot text chat pod, or
Michael McGarvey
16:28 – 16:31
Have you so that you get the same experience the same interaction?
Michael McGarvey
16:31 – 16:32
No matter what channel you consume?
Michael McGarvey
16:33 – 16:34
Yeah.
Michael McGarvey
16:34 – 16:34
Absolutely.
Michael McGarvey
16:34 – 16:37
I think that’s a big thing as well of that consistency.
Ryan Lester
16:37 – 16:52
And also there’s the value of the more day to you’re feeding into the system the better it’s going to get so you want those other data streams are those other sources of customer interaction or user interaction to be feeding the system say I totally agree.
Michael McGarvey
16:52 – 17:02
It’s almost like a breath of it’s a new life for breathing new life into some of these Legacy channels or another thing that I know you and I talked about is just some you highlight a couple of use cases.
Ryan Lester
17:02 – 17:19
So like onboarding of employees, I guess where else are you seeing interesting emerging use cases around Ai and chat Bots and kind of this, you know places where maybe unexpected or someplace where the you’re seeing companies leverage this technology in workflows.
Michael McGarvey
17:19 – 17:27
Well, I think part of it, you know, just just give you more of a we’ll start with the mundane one is that you know when we have an asset in a service and
Michael McGarvey
17:28 – 17:30
Internal support use case.
Michael McGarvey
17:30 – 17:37
We generally don’t like to put you know, big flashing notices of outages because generally they might be impactful to like one or two percent of the population.
Michael McGarvey
17:37 – 17:47
But again going back to that personalization thing finding ways of detecting when things happening and we start the interaction you can say Hey listen, I see you have access to this application.
Michael McGarvey
17:47 – 17:49
It’s currently down.
Michael McGarvey
17:49 – 17:56
Would you like me to notify you when it’s done and then finding a way of having that chat bot reach back out proactively saying, hey you listen your applications.
Michael McGarvey
17:56 – 18:09
Your application that was down is that be restored and then kind of offering that that bi-directional Communications way we is any way we can so that’s that’s one interesting way things like that.
Michael McGarvey
18:09 – 18:10
Yeah.
Michael McGarvey
18:10 – 18:14
So another thing I like about that because I like these parallel examples different different Industries.
Ryan Lester
18:15 – 18:23
We’ve had similar conversations even around like travel and Hospitality where you know, you have something like a weather event and I think about myself.
Ryan Lester
18:23 – 18:27
I’m a frequent traveler, you know for me, it’s brutal when you’re sitting in the airport and you’re
Ryan Lester
18:28 – 18:30
Well, where is my plane that’s coming here.
Ryan Lester
18:30 – 18:34
And is it delayed or is it even taking off or what’s its status?
Ryan Lester
18:34 – 18:37
So there’s all this kind of the information lives in systems.
Ryan Lester
18:37 – 18:43
I mean the airline, you know United knows where the plane is and they know what the weather is going on there.
Ryan Lester
18:43 – 18:45
So that’s like our well, where am I?
Ryan Lester
18:45 – 18:51
Oh my mobile device a my laptop and what are my preferences as a user and how much information do I want to do?
Ryan Lester
18:51 – 18:55
I want a lot of detail or a my casual travel or not really want to be bothered.
Ryan Lester
18:55 – 19:22
I rather just watch my movie in the in the airport and I think there’s all these great opportunities of once again making that data accessible to the user in a way they want to consume it and then to your point doing it in the right channel on complex systems become put systems and are actually just a bunch of discrete tasks has a plane that loaded is the weather clearing, you know is is the expectations can take off in the next ten minutes and then figure out what’s the right way to get that information to me as a user.
Ryan Lester
19:22 – 19:27
So I think that’s a great example of you know up time delivery.
Ryan Lester
19:27 – 19:28
I mean, there’s all these
Ryan Lester
19:28 – 19:35
Common frequent interactions kind of processes that really can be Amplified from a user perspective through something like AI.
Michael McGarvey
19:35 – 20:22
Yeah, and I think AI gives us an opportunity to perform these complex requests in a very simple way right you just you just talked about front finding that your flight status but be able to you know perform any kind of complex look up of information or or execution by, you know, giving a natural language command via voice or be a text message into a chatbot is really one of the powerful things where you can break down what somebody’s trying to do, even if they don’t give the the complete intent they misspelled words mispronounce words and being able to rearrange those things and figure out what they’re looking to do instead of having to go through the the way of hunting and pecking through different systems and find out the proper way just kind of give them the joke the gist of it.
Michael McGarvey
20:22 – 20:26
If you have to ask a few follow-up questions to get it right way.
Michael McGarvey
20:26 – 20:27
I can kind of put that together.
Michael McGarvey
20:28 – 20:29
Yeah, I agree.
Michael McGarvey
20:29 – 20:35
I mean that’s like the clock that clarification part of a conversation experience is really powerful where it’s you.
Ryan Lester
20:35 – 20:38
Don’t get lost in a branch of a tree where you’re kind of digging digging digging.
Ryan Lester
20:38 – 20:39
You’re like wait a minute.
Ryan Lester
20:39 – 20:39
Now.
Ryan Lester
20:39 – 20:49
I’m five layers deep in this is not at all what I was looking for, you know the conversationally I can help steer you and ask you those qualifying questions to once again help you get to that better outcome.
Michael McGarvey
20:49 – 20:57
So I totally agree that that’s that is really quite a bit transformational Yeah, and and understand the context between intense, right?
Michael McGarvey
20:57 – 21:04
So if you’re asking about one thing and then you ask him and related question be able to tie the two together so that you’re not asking all those qualifying questions again.
Michael McGarvey
21:04 – 21:05
Yeah.
Michael McGarvey
21:05 – 21:05
Absolutely.
Michael McGarvey
21:05 – 21:11
Yeah, and so we’ve touched on this a bit, but I want to dig in a bit more just based on your experience.
Ryan Lester
21:11 – 21:18
So, you know, there’s so there’s an agent population that these agents, you know, they work in contact centers.
Ryan Lester
21:18 – 21:28
They’re supporting a lot of these workflows today and we get a lot of questions around, you know AI is going to replace all the agents and everything is going to be powered by a I and I think you know, we’ve set the stage.
Ryan Lester
21:28 – 21:34
You say a lot of this is complementary to your agents, but there certainly will be an impact to the contact center.
Ryan Lester
21:34 – 21:41
So I want to dig in more their of kind of as you’re looking at today what you’re doing and then over the next 18 months two years.
Ryan Lester
21:42 – 21:47
What is going to be the weather kind of the big impacts you see on the contact center with things like a eye.
Michael McGarvey
21:47 – 21:52
Well, I think ultimately it’s a journey of trying to shift left.
Michael McGarvey
21:52 – 21:54
And I know that’s it’s overused a lot of times but it’s true.
Michael McGarvey
21:54 – 21:54
Right.
Michael McGarvey
21:54 – 22:01
So we’re trying to shift from higher skilled resources to lower skilled resources AI gives us an opportunity to do such a thing.
Michael McGarvey
22:01 – 22:17
So, you know, our contact center is generally handle High volumes of low quality interactions, meaning that they’re simple activities that sometimes user can do themselves or it simple knowledge and by shifting left and having it done through different channels like chat or even better with chatbot.
Michael McGarvey
22:17 – 22:27
We can free up those those those agents those operators to do more complex things and combining things whether it’s AI assisting the end user or AI.
Michael McGarvey
22:28 – 22:59
The operator we see very interesting things even some things where you know, there’s there’s some recent announcements for some cloud-based contact centers that allow you to stream in the audio real time and be able to drive the intent and then layering on AI to see what the agent and the operator are saying on a phone call and being able to derive things that they can do in an automated way or drive compliance is interesting application for AI between bought and and voice.
Michael McGarvey
22:59 – 23:23
So we see sort of AI allowing us to take more complex workloads bringing it down to the first level so they don’t have to be an expert on these things and they can get fed relevant information and be able to do those those robotic actions in the back end where you very viscous simplified a level 2 activity or a level 3 activity something a level one can do so as things get automated out.
Michael McGarvey
23:23 – 23:28
We have more activities being brought down so we don’t necessarily it as automating their
Michael McGarvey
23:28 – 23:33
Jobs out but really allowing them to focus on Lower quantity higher quality activities.
Michael McGarvey
23:34 – 23:58
Yeah, and I think there’s an interesting blurring of both like that kind of tech stack of like where you’re doing processing of what and you know, what lives as a voice data packet versus a text based data packet that and where it’s being processed and then also the kind of blurring of how you break that workload then down to an agent.
Ryan Lester
23:58 – 24:00
We’re like, what’s the agent doing in that workload?
Ryan Lester
24:00 – 24:03
And then what’s the AI doing and how are they complementing each other?
Ryan Lester
24:03 – 24:27
And so I think you brought up a number of good topics of there’s this interesting kind of evolution or very fast evolution is happening with like how people looking at ivr had looking at the context Center had looking at voice because voices become a blurred term between like voice assistant, you know home voice text to speech or speech to text.
Ryan Lester
24:28 – 24:50
Ivy are like all you know, all that stuff is becoming more and more blurry through these services and capabilities and I think it’s interesting of you know, how do you make the agent more effective more impactful and then and what are the things you want to push to AI for my workload perspective based on that Tech stack transforming so much over the next year or two years excetera.
Michael McGarvey
24:50 – 25:01
And for me one of the real challenges is we have multiple channels and we will probably always have multiple channels will have a voice Channel have a chat channel will have other channels video channels.
Michael McGarvey
25:01 – 25:08
But how do we homogenize the the back Nai of this right as a better service provider?
Michael McGarvey
25:08 – 25:16
I don’t want to have to develop multiple AI engines and and repositories and systems to service my customers.
Michael McGarvey
25:16 – 25:28
I’d rather have one that can service them and all channels right because then I can just spend my time on you know, refining intense refining the knowledge making sure it’s relevant.
Michael McGarvey
25:28 – 25:34
Vent working on our PA so they can be consumed by all these different despair channels through one day.
Michael McGarvey
25:34 – 25:36
I yeah, absolutely.
Michael McGarvey
25:36 – 25:38
I mean II totally agree.
Ryan Lester
25:38 – 25:49
I think that to me it becomes more of an eye, you know, I don’t think this is the best analogy but just of it becomes more of like a toolbox of like I need a tool.
Ryan Lester
25:49 – 26:01
I have my AI customer service tool that is really going to become awesome at understanding intense around those services or those customer interactions and then about other tools in my toolbox and it’s going to come down to okay.
Ryan Lester
26:01 – 26:03
What’s the best channel to engage?
Ryan Lester
26:03 – 26:04
Let’s figure that out.
Ryan Lester
26:04 – 26:09
But then once we engage, let’s get that into our intent engine to really nail.
Ryan Lester
26:09 – 26:10
What’s the thing?
Ryan Lester
26:10 – 26:10
What’s the question?
Ryan Lester
26:10 – 26:12
They’re trying to get answered or the problem.
Ryan Lester
26:12 – 26:17
They’re trying to solve and I think the more data you feed into that the better because it’s all gonna get better over time.
Ryan Lester
26:17 – 26:22
And it’s also going to give you a better sense of what are the most frequent intense coming in.
Ryan Lester
26:22 – 26:27
So, how are we going to let’s focus their first we’re going to things like our PA are further automation.
Ryan Lester
26:28 – 26:39
Because we’re going after those, you know, most painful or most used use cases and the more we can automate those the more freeing up our agents and more being operationally efficient Etc.
Ryan Lester
26:39 – 26:41
So I totally agree with that approach.
Michael McGarvey
26:41 – 26:57
I think for some people they look at more of there’s different tech stocks were doing parts of the business where siloed and they’re taking it more of like a one-off approach and I think that can certainly be dangerous it can be and I think this is still very merging area.
Michael McGarvey
26:57 – 27:00
I don’t think there are too many AI products on the market.
Michael McGarvey
27:00 – 27:01
They can’t really pull them together.
Michael McGarvey
27:01 – 27:10
But I think this is an area for maturity for this this industry right trying to kind of homogenize the different channels and having one common AI on the back end.
Michael McGarvey
27:10 – 27:25
So you’re right a lot of thought of teams might be structurally set up so that the The Voice team isn’t talking to the chat team isn’t talking to the artificial intelligent team and that’s a shame right having a having a flat structure where they’re all kind of art.
Michael McGarvey
27:25 – 27:28
We’re talking with her and users whether they be internal support.
Michael McGarvey
27:28 – 27:30
Or and a b2c relationship.
Michael McGarvey
27:30 – 27:36
How do we interface and Delight our customers using one common look and feel right or or sound and feel if you will?
Michael McGarvey
27:37 – 27:46
Yeah, I agree and I think the challenge, you know, it’s classic organizational structure challenge of just you know, the contact center has been using certain software for many years.
Ryan Lester
27:46 – 27:47
They’re comfortable with a certain model.
Ryan Lester
27:47 – 27:51
They also have to you know, run they have to be up the operational.
Ryan Lester
27:51 – 28:12
So every time they try to insert new technology adds risk initially, but can obviously add more value and and you know that organization, you know aren’t a experts to start so they may lie on an innovation group, but the Innovation group may have a broader AI initiative that’s you know, let’s use AI for you know, broader operational execution or supply chain or other things.
Ryan Lester
28:12 – 28:27
So there is this interesting Dynamic we’re seeing between kind of leveraging innovation group to help articulate the knowledge of the technology, but then making sure that where you’re investing in AI is relevant to your business use case.
Ryan Lester
28:28 – 28:38
And that’s something that you can invest in in the long term versus you know, you build something yourself and in reality of finding that there’s better Solutions in Market that are more tailored towards your use case.
Michael McGarvey
28:38 – 28:43
No, I would agree with that and I think you know in the chat area.
Michael McGarvey
28:43 – 28:56
It’s been sort of pretty open for a while and then the you know, ivr space it’s been very close but I think there’s been a number of products have come to market the past two years cloud-based contact centers that really give you an openstack to do that integration.
Michael McGarvey
28:56 – 29:02
Just people to spend the time and effort to kind of to do that integration of the back ends together.
Michael McGarvey
29:03 – 29:03
Yeah.
Michael McGarvey
29:03 – 29:06
Absolutely and I think so one good thing as well.
Ryan Lester
29:06 – 29:19
I’d say is to your point that more modern software taking them more open approach and I think the other nice thing is that there is some opportunity to leverage newer channels of you haven’t invested in chat or you’re looking at something like messaging.
Ryan Lester
29:19 – 29:27
So you’re like, hey we want to do something with a messaging Channel like, you know Microsoft teams for internal or Facebook Messenger.
Ryan Lester
29:28 – 29:36
You start to test their and start to build out your stack and then you can then leverage or for example in a mobile app and then you can say okay.
Ryan Lester
29:36 – 29:40
Now, let’s take those learnings and then apply them towards the bigger machine our contact center.
Ryan Lester
29:40 – 29:41
So there are opportunities.
Ryan Lester
29:41 – 29:52
I think the test and learn but but certainly I think that from a software perspective things have gotten more open to knowing that there’s maybe more than one ways companies are going to try to solve this.
Michael McGarvey
29:52 – 29:54
No Grabbers, huh?
Michael McGarvey
29:54 – 29:58
So I guess in you know as we’re kind of wrapping up here.
Ryan Lester
29:58 – 30:17
Is there any other advice or things as you’re kind of looking at you know in the work you’ve done we talked a little about this Evolution any other things that you’d say our best practices or things that companies should think about before maybe they’re diving into the Waters of AI or if they’re in the journey things that they should make sure they watch for.
Michael McGarvey
30:17 – 30:19
Yeah.
Michael McGarvey
30:19 – 30:27
So I mean I think having a clear strategy up front of exactly what you’re trying to accomplish is critical having the right mindset on.
Michael McGarvey
30:28 – 30:47
Let’s management is because if that’s the sort of certain the maturity level right now spending a lot of time making sure your knowledge is clean enough so it could be presented through Ai and doesn’t look like a huge page of like a wall of text super important having that that that those channels established and then layering on AI on top of it.
Michael McGarvey
30:47 – 31:27
That way you don’t have to drive volume to let’s say a new channel, right and that’s always difficult having a proven Channel starting in the layering on a on top of it is a best practice of course, and and those are those are the starting point and then I think beyond that having your your development teams have the right skill sets so that they can do Integrations with some of these third-party products to complete things like, you know, our PA and eventually work for things like integrate with voice assistants like Alexa for business is important, right just having folks that can they can do, you know, sir serverless Computing can do intend
Michael McGarvey
31:28 – 31:33
And that’s always very helpful as well to kind of make sure that you can be flexible with the system’s you want to integrate with.
Michael McGarvey
31:33 – 31:35
Yeah, that’s great.
Ryan Lester
31:35 – 31:50
And I think the two things you is a common theme in this podcast, but just a two big themes there is like understand your business objective and then of the right stakeholders involved that can help to adapt that as you scale drive more value.
Ryan Lester
31:50 – 31:54
So it’s like that knowing where you want to go and then taking steps to get there.
Ryan Lester
31:54 – 32:02
But then also having those internal teams that can kind of navigate organizationally navigate technically so that when you say, okay, we’re ready now to go integrate this with this next system.
Ryan Lester
32:02 – 32:04
Well, what does that mean has the data live in that system.
Ryan Lester
32:05 – 32:09
What’s our API who needs to be involved either.
Ryan Lester
32:09 – 32:13
There’s just as much a tech part of this as there is an organizational process part of it.
Michael McGarvey
32:13 – 32:17
agreed well, great.
Michael McGarvey
32:17 – 32:19
Well, I certainly appreciate the time.
Ryan Lester
32:19 – 32:24
This is a great conversation, and it’s interesting as I said earlier just to get your perspective on this because you work with so many companies.
Ryan Lester
32:24 – 32:27
I certainly won’t have you back again.
Ryan Lester
32:27 – 32:35
I think it’d be great to talk in a year from now and as we’re heading into, you know, the next wave of this where companies are going to go from there.
Ryan Lester
32:36 – 32:37
But thank you so much for the time.
Ryan Lester
32:37 – 32:41
Certainly appreciate the conversation and looking forward to the next time we can connect.
Ryan Lester
32:41 – 32:42
Yeah, you too.
Michael McGarvey
32:42 – 32:42
Thanks, Ryan.
Michael McGarvey
32:43 – 32:43
All right.
Michael McGarvey
32:43 – 32:44
Thanks so much.
Michael McGarvey
32:44 – 32:44
Have a great day.
Ryan Lester
32:44 – 32:45
You too.

Discuss / Read Comments

Leave a Reply

Leave a Reply

Explore our other AI Insights or Recent Posts.

Subscribe to the Bold360 Blog

Submit your email to get updates right to your inbox.