5 things to consider when building chatbots

1
00:00:16,170 –> 00:00:19,690
Good morning, evening and afternoon everyone. My name is Josh

2
00:00:19,690 –> 00:00:22,270
Reed from the global digital events team here at Genesys

3
00:00:22,270 –> 00:00:24,910
and I’ll be the moderator for today’s webcast. And I’ll

4
00:00:24,910 –> 00:00:27,900
start by saying welcome and thank you all for joining

5
00:00:27,900 –> 00:00:33,240
today’s webcast, five things to consider when building chatbots. So

6
00:00:33,340 –> 00:00:36,050
to ensure that you have the best experience viewing this

7
00:00:36,050 –> 00:00:38,870
webcast, I’m going cover a few housekeeping items before we

8
00:00:38,870 –> 00:00:42,610
get started. First, if you experience any problems while listening to today’s

9
00:00:42,610 –> 00:00:45,140
webcast, refresh your browser and make sure that it’s up

10
00:00:45,140 –> 00:00:48,330
to date to support HTML5, it may also help to

11
00:00:48,330 –> 00:00:50,970
switch over to something like Chrome or Mozilla Firefox as

12
00:00:50,970 –> 00:00:53,310
well as these are the best browsers that support the

13
00:00:53,310 –> 00:00:56,980
webcast platform. And if you’re having trouble seeing the slide

14
00:00:56,980 –> 00:00:59,640
window, you can actually enlarge that window by dragging the

15
00:00:59,640 –> 00:01:04,760
bottom right corner. Also, please know that this webcast is designed

16
00:01:04,760 –> 00:01:07,970
to be an interactive experience between you and our presenters so

17
00:01:07,970 –> 00:01:10,459
at any time during this webcast, feel free to submit

18
00:01:10,459 –> 00:01:13,300
those questions in the Q& A window in the top middle and

19
00:01:13,300 –> 00:01:15,970
we’ll answer as many as we can throughout the presentation

20
00:01:16,330 –> 00:01:18,700
and at the end with our live Q& A session.

21
00:01:18,870 –> 00:01:21,850
However, if time does get away from us and we aren’t

22
00:01:21,980 –> 00:01:24,800
able to answer your question in the live setting, don’t

23
00:01:24,800 –> 00:01:27,670
worry we will follow up with you via email after

24
00:01:27,670 –> 00:01:31,680
the webcast conclusion within the next few business days. And

25
00:01:31,680 –> 00:01:34,380
know that this webcast is being recorded, so if you miss

26
00:01:34,380 –> 00:01:39,180
anything, have to refresh or want to have playback capabilities later,

27
00:01:39,180 –> 00:01:42,240
you will receive the on- demand recording via email from

28
00:01:42,240 –> 00:01:44,850
me within the next few business days so just be

29
00:01:44,850 –> 00:01:49,460
on the lookout for that. And also, at any time

30
00:01:49,460 –> 00:01:51,770
during the webcast, feel free to check out the resource

31
00:01:51,770 –> 00:01:54,910
box below the Q& A window. Clicking through won’t take you

32
00:01:54,910 –> 00:01:57,010
away so don’t worry about that, it’ll open up on a

33
00:01:57,010 –> 00:02:01,410
new tab, but these resources will really expand on today’s topic.

34
00:02:02,850 –> 00:02:05,930
And lastly, we welcome and appreciate your feedback. So you’ll

35
00:02:05,930 –> 00:02:07,780
have the opportunity to fill out a short survey on

36
00:02:07,780 –> 00:02:12,020
today’s presentation. The survey’s at the last icon below or it’ll just

37
00:02:12,130 –> 00:02:14,010
show up automatically at the end if you don’t want to worry about

38
00:02:14,120 –> 00:02:19,130
it yet. Okay, I promise. That’s all of the housekeeping stuff.

39
00:02:19,810 –> 00:02:23,139
Today we have two excellent presenters excited to discuss how

40
00:02:23,139 –> 00:02:27,760
to create chatbots that improve customer experience instead of increasing

41
00:02:27,760 –> 00:02:31,400
customer frustration. And we will give you an inside look

42
00:02:31,400 –> 00:02:36,100
on how Bosch Service Solutions integrated multiple natural language understanding

43
00:02:36,100 –> 00:02:42,450
or NLUs technologies, within a single platform. I’m thrilled to

44
00:02:42,450 –> 00:02:47,530
introduce Torsten Moritz the service owner of communications systems with

45
00:02:47,530 –> 00:02:51,500
Bosch Service Solutions as well as Genesys’s very own Joe

46
00:02:51,500 –> 00:02:55,110
Schufo, the product marketing director. So with all that being

47
00:02:55,110 –> 00:02:57,010
said, I’m actually going to go ahead and hand things

48
00:02:57,010 –> 00:02:59,970
off to the men of the hour here, Torsten, the floor is yours. Thanks

49
00:03:03,770 –> 00:03:07,550
a lot Josh. So welcome everybody. So maybe a short

50
00:03:07,550 –> 00:03:12,610
introduction, not about Bosch, I think you all are familiar with the company, most

51
00:03:12,610 –> 00:03:17,030
of you will have some dishwashers, other devices at your

52
00:03:17,030 –> 00:03:30,600
house. So Bosch is kind of a very huge company with multiple divisions. This is one of the reasons or one of the backgrounds we need

53
00:03:30,600 –> 00:03:55,950
to understand for our use cases. So one of the interesting story of Bosch is the amount of employees and what I outlined also, the amount of divisions we have. As it’s on the first slide, I’m belonging to Bosch Service Solutions, which is an owned division. We are doing business process outsourcing, so we are on BPO provider for

54
00:03:55,950 –> 00:04:02,280
internal and external customers. So I wanted to give you the framework

55
00:04:02,650 –> 00:04:06,180
where we are working in and also how the Genesys platform fits

56
00:04:06,180 –> 00:04:12,930
into our business. So you’re all familiar with Bosch so let’s move

57
00:04:12,930 –> 00:04:32,720
over to my division, Bosch Service Solutions. I will then go over to the next slide, so Bosch Service Solutions started out of a called devices were delivered by Bosch and the missing part was the customer service. And out of this distribution become an a BPO provider, so

58
00:04:46,740 –> 00:04:50,290
the footprint is quite large. So the variety of different customers

59
00:04:50,290 –> 00:04:57,370
we have is quite big. So different countries, different divisions,

60
00:04:57,460 –> 00:05:07,210
different areas of consumer services, so that lead to a huge demand of diversity in our portfolio for our chatbots and

61
00:05:07,210 –> 00:05:37,110
other services we offer with the Genesys platform. So looking over to the different services we offer at our division, the area where we focus now with this webcast is the customer experience services. As you see different deliveries and they really, at the end, most likely covering

62
00:05:37,110 –> 00:05:43,160
all of your businesses, so we’re doing a different business tenure but at

63
00:05:43,160 –> 00:05:46,610
the end, I think we’re not BPO but we are

64
00:05:46,610 –> 00:05:49,990
covering most likely from the content, one of your businesses

65
00:05:50,070 –> 00:05:54,560
where you’re from. So at the end, the things I’m

66
00:05:54,970 –> 00:05:58,070
trying to outline today are more general but it should

67
00:05:58,070 –> 00:06:01,080
be able for you to pick up or to reuse

68
00:06:01,170 –> 00:06:04,390
when you try to integrate both and be successful with this.

69
00:06:05,500 –> 00:06:10,450
One of the major topics next to the variety of areas

70
00:06:10,950 –> 00:06:15,610
is the amount of locations we’re operating the Genesys platforms.

71
00:06:17,210 –> 00:06:19,700
On the bottom you can see a different location from

72
00:06:19,700 –> 00:06:24,450
our division and on 15 locations we’re running the Genesys

73
00:06:24,450 –> 00:06:29,250
platforms. So out of this we have a huge demand of

74
00:06:30,080 –> 00:06:36,529
different requirements when it comes to legal requirements, local requirements

75
00:06:36,529 –> 00:07:07,520
and so on. The combination of these different source of have a huge impact on how we build up our platform and how we’re pushed to deliver new functionalities and also adapt to the market. The name in our terms for the Genesys platform is Unified Communications platform and as I mentioned, 15 different locations all connected to one single platform and because of this customer demand, we’re doing a lot of integrations to what’s difference systems and also in distribution case, also different bot platforms. So I think that’s kind

76
00:07:07,570 –> 00:07:39,290
of unique, most likely you will start with one bot and then try to build it up for your customers for this bot. In our case, we did multiple ones. This leads me to

77
00:07:39,290 –> 00:07:48,820
our first question, our first poll question. Interesting part for us would be how many Bot

78
00:07:48,820 –> 00:07:54,370
technology companies are you evaluating? So please give your vote

79
00:07:54,720 –> 00:08:02,130
and we will try to focus on your answers- And

80
00:08:02,130 –> 00:08:08,080
while we’re- to say, Torsten, I think this is a

81
00:08:08,080 –> 00:08:10,690
really interesting question while we’re waiting to see how many

82
00:08:10,690 –> 00:08:12,970
people are looking at different vendors just like you are

83
00:08:12,970 –> 00:08:24,070
as well. Exactly. Because of our business, we are even sometimes asked

84
00:08:24,200 –> 00:08:28,410
by a customer to integrate a specific technology, not only

85
00:08:28,810 –> 00:08:35,520
in the Bot area, so also other or AI technologies. They

86
00:08:35,520 –> 00:08:39,360
already decided in their company for some specific window and

87
00:08:39,360 –> 00:08:42,240
they are more or less asking us to adapt to

88
00:08:42,240 –> 00:08:44,630
this one with our part of the customer service we

89
00:08:44,630 –> 00:08:50,390
are delivering for them. I want to give you some more time

90
00:08:50,390 –> 00:09:06,280
to vote, we have quite some amount of attendees. Josh,

91
00:09:06,280 –> 00:09:16,580
I think we’re good to go. Yep. I’ll go and push those results now. Okay. So as

92
00:09:16,580 –> 00:09:20,500
you see, the majority is having a 0- 2, which most

93
00:09:20,500 –> 00:09:22,850
likely would be, I think I would consider a one.

94
00:09:24,820 –> 00:09:31,130
Actually, we are already integrated three different windows, one of

95
00:09:31,240 –> 00:09:43,370
the reasons as I said is we had a different demand from the market, and also we’re pushed into different directions due to strategy

96
00:09:43,880 –> 00:09:50,100
decisions outside of our organization. I’m interested to see that

97
00:09:50,390 –> 00:09:56,550
people are investigated more than six, so that’s kind of a huge amount

98
00:09:56,550 –> 00:10:01,840
of effort to put into really, but at the end I think really well

99
00:10:01,840 –> 00:10:12,140
invested time to find out what happens. So the amount

100
00:10:12,140 –> 00:10:15,000
of effort you put in at the end could give

101
00:10:15,000 –> 00:10:20,980
you really beneficial results, so the interesting part for us

102
00:10:20,980 –> 00:10:27,790
was that we could easily integrate the different so we started

103
00:10:27,790 –> 00:10:34,320
with Inventor, then the next integration we did with Google Dialogflow and the recent

104
00:10:34,850 –> 00:10:40,790
integration we did with a smaller German company, there were different reasons

105
00:10:41,170 –> 00:10:45,740
and different starting points why we did it, and one

106
00:10:45,740 –> 00:10:49,950
was kind of internal evaluation towards Inventor, then there was kind of

107
00:10:50,530 –> 00:10:53,440
a trend to go over with Google Dialogflow and the

108
00:10:53,440 –> 00:10:57,330
last one was because we could also run this on Premise.

109
00:10:57,420 –> 00:11:04,410
So for all of these three we were not starting with the Bot

110
00:11:04,410 –> 00:11:08,470
capabilities and so we always started how easy it is

111
00:11:09,420 –> 00:11:18,530
to provide this so called and this most likely implies that they

112
00:11:18,530 –> 00:11:22,179
think that’s kind of the exception. For us it was always the other

113
00:11:22,179 –> 00:11:25,100
way around, we always had the Bot only as an

114
00:11:25,100 –> 00:11:28,740
extension to the customer service. So we didn’t see what

115
00:11:29,970 –> 00:11:33,350
steps our agent kicks in as an exception. So for us

116
00:11:33,600 –> 00:11:37,380
we’ve lost the standard we need to cover with any Bot integration even before the

117
00:11:37,380 –> 00:11:43,010
Genesys Bot Gateway was available. So from the beginning we focused really on

118
00:11:43,760 –> 00:11:48,110
how the customer can be transferred properly to the agent without

119
00:11:49,120 –> 00:11:51,970
having a break in the customer journey or the agent

120
00:11:51,970 –> 00:11:56,880
starting from scratch. So a very interesting situation I had when

121
00:11:56,880 –> 00:11:59,720
I was at a customer meeting and the Bot topic

122
00:11:59,720 –> 00:12:07,610
came up and they told me they already did some Bot testing

123
00:12:07,610 –> 00:12:12,360
introduction and they badly failed. And I asked them how easy

124
00:12:12,360 –> 00:12:16,760
was it for the customer to reach the agent? And my intention behind

125
00:12:16,760 –> 00:12:20,340
the question was how many loops did the customer do?

126
00:12:20,420 –> 00:12:22,940
So how many? I did not understand you, please repeat

127
00:12:24,110 –> 00:12:32,270
cycles. And they told me that there was no agent involved at all, so no agent handover at all. So at the end they

128
00:12:32,920 –> 00:12:38,440
tried to start with a fresh Bot without considering any fallback.

129
00:12:39,290 –> 00:13:04,030
Especially if you start into production with a new bot, the chances for these other amount of these agent handovers or fallbacks are quite high and Also, I think one of the problems that most people see it as an issue and the expectations are way too high especially from management or from the outside towards the Bot project, they call it this. The

130
00:13:04,030 –> 00:13:17,240
scope management and also the expectation management, especially for the rollout, that’s really

131
00:13:17,240 –> 00:13:23,890
critical. So designing the Bot in the first integration to

132
00:13:23,890 –> 00:13:31,320
really also be fine was delivering this customer experience, which is

133
00:13:31,320 –> 00:13:35,910
not broken by Bot steps out, agent comes in, that’s already

134
00:13:35,940 –> 00:13:40,730
a good achievement. So for the customer there is no

135
00:13:41,290 –> 00:13:46,890
disadvantage. That’s the bottom line you should be able to reach at

136
00:13:46,890 –> 00:13:52,300
anytime for any project. If you identify that because of

137
00:13:52,300 –> 00:14:10,410
the nature of your service you will not be able to 100% self service the customer with the bot, that’s also totally fine. The qualification or classification approach for a Bot

138
00:14:10,410 –> 00:14:20,320
is also valuable to a service. So the expectation management and also how you really use the Bot and integrate

139
00:14:20,560 –> 00:14:23,490
it in your service is another thing that’s really a key

140
00:14:23,490 –> 00:14:31,220
to success. So trying to achieve from the beginning this 100% self service buy Bot for all

141
00:14:31,220 –> 00:14:34,300
use cases, I have my full customer service that will

142
00:14:34,300 –> 00:14:37,580
not, that’s obvious will not work and also that’s not

143
00:14:37,810 –> 00:14:40,470
a goal which you should try to achieve because there

144
00:14:40,470 –> 00:14:43,970
will be any way specific customers will use where you

145
00:14:43,970 –> 00:14:47,980
will never be able to meet their requirements. One of

146
00:14:47,980 –> 00:14:51,950
the other trends we saw in projects and discussions that

147
00:14:52,270 –> 00:14:56,020
more or less the full business logic, it moves into

148
00:14:56,020 –> 00:15:00,650
the bot. So that’s something which I can understand because

149
00:15:00,760 –> 00:15:03,850
now different teams will handle the Bot next to the

150
00:15:03,850 –> 00:15:07,800
Genesys developers and so on, but this is kind of

151
00:15:08,630 –> 00:15:15,550
wasting some resources. So the trend of moving everything out of the routing

152
00:15:16,220 –> 00:15:43,820
kind of limits or reduces this possibilities. An example would be, I don’t want to offer the Bot for with customers. So if I don’t do this or backing integration in the routing, then I will engage a Bot to decide not to handle this conversation. So that’s kind of, especially when you look at the cost and license cost, it’s kind of a very expensive decision because in the routing, that’s more or less for free

153
00:15:43,820 –> 00:15:54,750
because you don’t license decisions there. That’s a different, that’s kind of easy move, let’s move everything into the Bot because there we

154
00:15:54,750 –> 00:15:58,990
can easily build and design our flows. That’s kind of contradicting

155
00:15:58,990 –> 00:16:02,830
from some of the service designs and also sometimes from

156
00:16:02,830 –> 00:16:09,500
the cost efficiency you try to gain. I said, the

157
00:16:09,500 –> 00:16:14,930
overall design, I think that’s critical, that’s also something Joe will show in his demo,

158
00:16:15,520 –> 00:16:17,960
I don’t have that many slides. So you will see

159
00:16:17,960 –> 00:16:22,380
a demo from Joe in a few minutes. I said

160
00:16:22,380 –> 00:16:26,340
that the expectation management, I think that’s really critical, so

161
00:16:26,340 –> 00:16:30,380
I had multiple discussions where really people were already afraid

162
00:16:30,400 –> 00:16:35,440
before the project that they cannot deliver the management expectations towards the

163
00:16:35,760 –> 00:16:40,360
bot. The culmination of, I need to build up really new

164
00:16:40,360 –> 00:16:45,530
teams for training the bot, designing all these use cases and

165
00:16:46,210 –> 00:16:51,880
building up this combination between maybe the IT for the, let’s

166
00:16:51,880 –> 00:16:54,730
call it classical Genesys stuff, and the new Bot I need to

167
00:16:54,730 –> 00:16:59,030
integrate, that’s kind of an already challenging part to find out

168
00:16:59,080 –> 00:17:05,250
the responsibilities and also to identify where to focus and

169
00:17:05,970 –> 00:17:13,960
which content to do where. So the move from routing

170
00:17:14,130 –> 00:17:17,760
towards Bot could have some beneficial parts, but at the

171
00:17:17,760 –> 00:17:22,790
end, especially looking at this full service design, that could be

172
00:17:23,180 –> 00:17:27,030
kind of a misleading trend. A lot of people are already trying

173
00:17:28,300 –> 00:17:32,400
to do because of their starting point but only without

174
00:17:32,400 –> 00:17:37,440
considering the routing. So this is next also to the

175
00:17:37,440 –> 00:17:44,040
reporting part, something where I would recommend to really focus on, to not

176
00:17:44,340 –> 00:17:48,690
focus on Bot only, see the full chain of the

177
00:17:49,160 –> 00:17:58,869
service. Other interesting things were that the exchange between Genesys

178
00:17:58,869 –> 00:18:02,440
and the bot, whatever it is, it’s also kind of powerful

179
00:18:03,060 –> 00:18:07,640
so if you already get some information from the website,

180
00:18:08,090 –> 00:18:11,780
you can use this to maybe call the Bot different, even

181
00:18:11,780 –> 00:18:15,780
engage with different bots. So maybe in some countries, you want

182
00:18:15,780 –> 00:18:19,740
to offer a cheaper Bot if you’re really cost driven

183
00:18:19,740 –> 00:18:23,250
in other countries you offer a more expensive bot. So

184
00:18:23,340 –> 00:18:28,650
one of the huge benefits we have seen is because the Bot integrated

185
00:18:28,650 –> 00:18:32,440
on the webpage itself, it’s integrated in general routing, we can

186
00:18:32,500 –> 00:18:36,800
easily kind of switch on, switch off the bots. So

187
00:18:37,540 –> 00:18:43,170
looking at the classical works split where most likely you

188
00:18:43,170 –> 00:18:46,460
don’t own your own or you’re not responsible for your

189
00:18:46,460 –> 00:18:50,200
corporate website, you need to go either somewhere internal or

190
00:18:50,200 –> 00:18:54,609
even to a agency to have something implemented in the webpage, this is

191
00:18:54,609 –> 00:19:00,010
really an interesting benefit to enable or disable the Bot and

192
00:19:00,010 –> 00:19:04,020
routing for a rollout scenario while not changing anything on

193
00:19:04,320 –> 00:19:11,690
the webpage. So only a webpage will still have your and if a Bot steps into

194
00:19:11,690 –> 00:19:15,940
this conversation, that’s fully up to the routing logic, so

195
00:19:16,980 –> 00:19:20,960
maybe based on the IP range, the customer is coming from,

196
00:19:20,960 –> 00:19:24,960
maybe on this page, I’m offering it or on this

197
00:19:24,960 –> 00:19:32,780
segment I’m offering the bot. So going back to the topic, really don’t waste the Genesys

198
00:19:32,830 –> 00:19:36,810
routing resources that we use them also for these decisions.

199
00:19:36,810 –> 00:19:41,950
So from a more marketing, general point of view, you

200
00:19:41,950 –> 00:19:44,970
would say treat the bots similar to the agents. So

201
00:19:44,970 –> 00:19:48,480
if you invest a lot of effort in your agent

202
00:19:48,480 –> 00:19:53,990
routing, why won’t you reuse this benefits also for your kind

203
00:19:53,990 –> 00:19:59,840
of Bot routing? This is kind of a different approach

204
00:20:00,220 –> 00:20:02,750
than other companies do where the Bot is in front

205
00:20:03,150 –> 00:20:08,470
and then they need to develop an agent kind of a

206
00:20:08,470 –> 00:20:27,609
help or something like this. We have kind of standard generals into actions, we have all the benefits to really reuse from the Genesys platform. And one of the very interesting benefits we have seen and

207
00:20:28,230 –> 00:20:31,910
also compared to other windows, which are coming from the

208
00:20:31,910 –> 00:20:35,890
Bot area and then trying to add some agent experience

209
00:20:35,890 –> 00:20:45,580
afterwards. The re- usability of a Genesys integrated chatbot is kind of

210
00:20:46,480 –> 00:20:50,020
very easy and simple at least for plain text to

211
00:20:50,020 –> 00:20:54,140
transfer from one to another channel, so if you’re doing a chat

212
00:20:55,310 –> 00:20:58,830
integration with Genesys or the Bot Gateway, you don’t need

213
00:20:58,830 –> 00:21:06,760
to change anything when you want to migrate this kind of integration in the towards a business

214
00:21:06,760 –> 00:21:15,640
chat or a Facebook private message. This is kind of a very powerful and very cost saving

215
00:21:15,640 –> 00:21:20,490
mechanism for us, where we are able to fulfill the market demand

216
00:21:20,990 –> 00:21:29,710
because of this, what I outlined at the beginning, this variety of different requirements and different use cases

217
00:21:30,140 –> 00:21:35,580
we have. We are also forced to deliver multiple channels and

218
00:21:35,580 –> 00:21:41,490
they can pop up quite frequently and quite immediately for

219
00:21:41,490 –> 00:21:46,460
us. So the benefit to at least start with the basic

220
00:21:46,460 –> 00:21:56,730
chat, text chat that’s really beneficial to achieve for us, so the

221
00:21:56,730 –> 00:21:58,670
overall reception is quite good for us and I think that’s all for now, I don’t want

222
00:21:58,670 –> 00:22:11,910
to waste too much time from Joe’s presentation. So as Josh already mentioned at the beginning,

223
00:22:12,500 –> 00:22:16,200
we are happy to take some questions and answer them and

224
00:22:16,200 –> 00:22:21,780
we have also some time left at the end, so

225
00:22:21,780 –> 00:22:27,040
let’s hand over to Joe, thanks. Awesome. Thanks so much

226
00:22:27,040 –> 00:22:30,869
Torsten. So as we just mentioned, I want to go

227
00:22:30,869 –> 00:22:33,869
through a demonstration with you, really hitting on five main

228
00:22:33,869 –> 00:22:37,359
points about building chatbots. But before I do that, I

229
00:22:37,359 –> 00:22:39,500
am nosy and I have some poll questions to keep

230
00:22:39,500 –> 00:22:42,570
you guys engaged here as well. So let’s get this

231
00:22:42,570 –> 00:22:45,440
poll questions, let’s go through the demo and then we’ll

232
00:22:45,440 –> 00:22:47,400
hit the Q&A at the very end of this so that we

233
00:22:47,400 –> 00:22:49,270
can get some of those questions I know you’ve got

234
00:22:49,270 –> 00:22:52,760
for Torsten. The first question I have for everyone here

235
00:22:52,760 –> 00:22:56,480
today is what types of solutions are you considering to

236
00:22:56,480 –> 00:23:00,050
help with your Bot initiative? And this is something we

237
00:23:00,050 –> 00:23:04,060
hit home with Torsten’s speech earlier. Sometimes we look at

238
00:23:04,060 –> 00:23:07,810
this in house, other times we’re reaching out to other

239
00:23:07,830 –> 00:23:12,930
NLU providers like Dialogflow from Google or even local vendors

240
00:23:12,930 –> 00:23:16,000
that can run on Premises like we mentioned today. And

241
00:23:16,000 –> 00:23:18,150
I’m really interested to see what we’re seeing from people

242
00:23:18,150 –> 00:23:22,790
out there. So I’m going to give this a few more minutes, by minutes

243
00:23:22,790 –> 00:23:25,990
I mean seconds, I’m not that patient. And let’s go

244
00:23:25,990 –> 00:23:29,400
ahead and see those results here. This is a good

245
00:23:29,400 –> 00:23:31,600
split, we see that a lot of people are looking

246
00:23:31,600 –> 00:23:34,820
at building some of these solutions in house, reconsidering a

247
00:23:34,820 –> 00:23:38,010
Bot initiative, some of them looking at system integrators and

248
00:23:38,010 –> 00:23:40,760
just as equally look at it from a software standpoint.

249
00:23:41,320 –> 00:23:44,440
This is important, this is something we hit on within

250
00:23:44,440 –> 00:23:48,730
this speech earlier, is having that flexibility. You can choose

251
00:23:48,730 –> 00:23:51,450
those vendors, you can choose to design some of these

252
00:23:51,450 –> 00:23:54,500
in house and you may even go with a integrator as well.

253
00:23:54,790 –> 00:23:59,470
Getting that ability to choose if, not when or if what’s

254
00:23:59,470 –> 00:24:03,300
possible, really does give that some power. I’ve got one

255
00:24:03,300 –> 00:24:05,140
more question for you before we take a look at

256
00:24:05,140 –> 00:24:08,859
that demo. And that question is, if you have deployed

257
00:24:09,010 –> 00:24:11,750
one or more bots, what type of problems are you

258
00:24:11,750 –> 00:24:16,160
having right now? Now, I realize that is a extremely

259
00:24:16,160 –> 00:24:19,030
pointed question, but what I’m trying to get to is, what are

260
00:24:19,369 –> 00:24:22,140
the pain points? What are the areas that you think you could use

261
00:24:22,140 –> 00:24:24,980
help when it comes to managing these bots? And we

262
00:24:24,980 –> 00:24:28,320
saw earlier that we’ve got a wide amount of you

263
00:24:28,320 –> 00:24:30,630
that are looking at anywhere from zero bots all the

264
00:24:30,630 –> 00:24:33,490
way up to six and more, so as we add

265
00:24:33,490 –> 00:24:40,280
more and more to the equation here, things change, complexities increase. Let’s give

266
00:24:40,280 –> 00:24:42,320
that one more second and take a look at those

267
00:24:42,320 –> 00:24:47,700
results here. I love this. As we start looking at

268
00:24:47,700 –> 00:24:50,619
more and more bots, the answer that we’re seeing overwhelmingly is

269
00:24:50,960 –> 00:24:53,380
we don’t know what we don’t know. And that’s a

270
00:24:53,380 –> 00:24:56,080
really interesting piece, one of the key components of what

271
00:24:56,080 –> 00:24:59,640
we work with is reporting and being able to understand

272
00:25:00,020 –> 00:25:03,609
what’s going on within those bots? Are they failing? Are they transferring

273
00:25:03,609 –> 00:25:06,380
to agents? And what type of impact are they having

274
00:25:06,570 –> 00:25:11,530
with those agents? so what I want to do here is

275
00:25:11,740 –> 00:25:16,160
jump into that demo, and we talk about demonstrations but I want

276
00:25:16,430 –> 00:25:18,980
to show, there’s a story here on how we can make

277
00:25:18,980 –> 00:25:21,850
that experience great for our customer and how we can make

278
00:25:21,850 –> 00:25:27,580
that experience really between a Bot and an agent. The five main points that I want

279
00:25:28,080 –> 00:25:31,070
to hit on today are knowing when to engage, and

280
00:25:31,070 –> 00:25:34,160
that is the very beginning of this. Using tools that

281
00:25:34,160 –> 00:25:37,490
we’ll see to jump in right at the right moment with

282
00:25:37,490 –> 00:25:41,660
full understanding of the context. So as we engage these

283
00:25:41,660 –> 00:25:44,970
customers, are we using that current journey that we’re aware

284
00:25:44,970 –> 00:25:47,890
of to actually drive the conversation that we have here

285
00:25:47,890 –> 00:25:52,680
today? We mentioned this in the speech earlier, but a key

286
00:25:52,680 –> 00:25:56,680
component is, do not limit channels. We talk about chatbots

287
00:25:56,680 –> 00:25:58,660
and I think we have a natural tendency to think that a

288
00:25:58,660 –> 00:26:03,940
chatbot is just on a website, but a chatbot is Facebook Messenger,

289
00:26:04,109 –> 00:26:08,000
chatbot is WhatsApp, a business chat, it could be SMS,

290
00:26:08,540 –> 00:26:10,920
maybe it’s not even a chatbot at all, maybe it’s a voicebot

291
00:26:10,920 –> 00:26:14,619
and it’s on a phone call. Don’t limit channels and

292
00:26:14,619 –> 00:26:18,230
customers shouldn’t have a poor experience simply because we were

293
00:26:18,230 –> 00:26:21,300
only able to provide a good experience in one of those channels.

294
00:26:22,290 –> 00:26:24,240
Two other key components I’ve really want to hit on,

295
00:26:24,240 –> 00:26:27,900
and Torsten brought up this really nicely, is make sure

296
00:26:28,230 –> 00:26:30,840
that agents are always kept in mind. We should be

297
00:26:30,840 –> 00:26:33,450
able to blend these conversations as we talk about the

298
00:26:33,450 –> 00:26:37,770
interactions today from a chatbot to an agent, with full

299
00:26:37,770 –> 00:26:44,320
context of what has happened. And lastly there are so many great

300
00:26:44,320 –> 00:26:47,640
tools out there. I know this is a Genesys webinar

301
00:26:47,640 –> 00:26:49,850
but we hear more and more about some of the

302
00:26:49,850 –> 00:26:52,660
great tools we’re seeing out of using Microsoft, out of

303
00:26:52,660 –> 00:26:56,500
using Google Dialogflow, even Amazon and you should be able to do

304
00:26:56,500 –> 00:27:01,260
that. This is absolutely a situation where great technology shouldn’t

305
00:27:01,260 –> 00:27:04,700
be unavailable to you and you should be choosing when not if

306
00:27:04,700 –> 00:27:07,440
to try all these other Bot technologies for your business

307
00:27:07,440 –> 00:27:11,680
objectives. So let’s get to the demo here. I’ve babbled

308
00:27:11,680 –> 00:27:13,560
quite a bit, it’s time I actually show you some

309
00:27:13,560 –> 00:27:16,380
cool things. What I wanted to do today was walk

310
00:27:16,380 –> 00:27:19,060
through a story, and we’ve got a fake insurance company

311
00:27:19,060 –> 00:27:22,660
here to really set the tone G Insurance. We’re going

312
00:27:22,760 –> 00:27:25,220
to take a look at that experience from the perspective

313
00:27:25,220 –> 00:27:29,160
of a customer and then I’ll also be playing the agent’s side.

314
00:27:29,970 –> 00:27:32,580
And it all starts about, what happens when a customer

315
00:27:32,580 –> 00:27:36,260
is trying to engage with you but you sometimes make

316
00:27:36,260 –> 00:27:39,369
it hard to find that information, are we losing them?

317
00:27:39,369 –> 00:27:42,820
And how do we get them back? Even beyond that,

318
00:27:43,210 –> 00:27:45,940
losing customers to a lack of support or maybe even

319
00:27:46,060 –> 00:27:49,619
a missed sales opportunity is not only frustrating, it’s easily

320
00:27:49,619 –> 00:27:52,960
preventable. So let’s go through some of those key components

321
00:27:52,960 –> 00:27:55,710
that we can use based off those five points to

322
00:27:55,710 –> 00:27:59,010
see how we could make this a great situation. Here I

323
00:27:59,010 –> 00:28:02,410
am, I’m on the G Insurance website and let’s say

324
00:28:02,410 –> 00:28:06,700
I’m an existing customer that potentially has home insurance. So

325
00:28:06,700 –> 00:28:10,310
often, we see emails going out to customers saying, if

326
00:28:10,310 –> 00:28:13,200
you’ve gotten home insurance, you could bundle that together with

327
00:28:13,210 –> 00:28:16,890
your car insurance and potentially save some money. Now, we’ve

328
00:28:16,890 –> 00:28:20,670
spent all of this time crafting these email messages, getting

329
00:28:20,670 –> 00:28:23,640
them to our website and going out of our way

330
00:28:23,640 –> 00:28:27,470
to drive that engagement. So here I am just logging

331
00:28:27,470 –> 00:28:30,020
into the website, and I want to point something out,

332
00:28:30,680 –> 00:28:34,940
that chatbots are quite often used as an excuse for bad

333
00:28:34,940 –> 00:28:38,400
web design. Now, I know that’s odd but think about

334
00:28:38,400 –> 00:28:41,040
that, we have these websites that make it hard to

335
00:28:41,040 –> 00:28:43,640
find the information we need. I don’t see anything about

336
00:28:43,640 –> 00:28:46,720
car insurance here that’s actually getting me for what I

337
00:28:46,720 –> 00:28:49,060
want to find. Maybe I’ll go to the get a

338
00:28:49,070 –> 00:28:52,940
quote page and even though it has information about me

339
00:28:52,940 –> 00:28:55,980
applying or in getting that quote, I’ve got some questions before I

340
00:28:55,980 –> 00:28:59,850
do that, so what happens? Well, if I can’t find

341
00:28:59,850 –> 00:29:04,160
the information I’m looking for, I might leave. So I’ve gone on

342
00:29:04,510 –> 00:29:06,940
the search bar and I’m looking at car insurance, it’s

343
00:29:06,940 –> 00:29:11,180
all about things that I’m not looking for before I apply. This is

344
00:29:11,180 –> 00:29:15,720
frustrating. You’ve gotten me to the page, I’m actively searching

345
00:29:15,720 –> 00:29:18,940
for information on car insurance and I can’t find the

346
00:29:18,940 –> 00:29:22,270
information I need. So what can we do to change that?

347
00:29:23,040 –> 00:29:25,230
Well, we’ve got a few pieces of technology on the

348
00:29:25,230 –> 00:29:30,220
backend that can engage with predictive chatbots, so knowing when

349
00:29:30,220 –> 00:29:33,270
to engage with the customer and why we should engage?

350
00:29:33,990 –> 00:29:35,650
And the way we can do that is with something

351
00:29:35,650 –> 00:29:40,960
called Genesys Altocloud. So Genesys Altocloud is a predictive engagement

352
00:29:40,960 –> 00:29:44,560
technology, and to show that, we can see a live

353
00:29:44,560 –> 00:29:47,540
now dashboard of every single customer that is on our

354
00:29:47,540 –> 00:29:52,610
website. As they’re visiting, what are they searching and did

355
00:29:52,610 –> 00:29:55,560
they match current segments? So when we think about customer

356
00:29:55,560 –> 00:29:59,300
journey, did we send them an email about a potential

357
00:29:59,300 –> 00:30:03,590
offer? If we reached out and said hey, Sarah, come

358
00:30:03,590 –> 00:30:05,500
check out what we have for you in terms of

359
00:30:05,500 –> 00:30:08,500
that car insurance information so that you could bundle that together,

360
00:30:09,320 –> 00:30:11,770
this is something we should keep in mind. And as

361
00:30:11,770 –> 00:30:14,370
we look at all these customers, we might even drive

362
00:30:14,370 –> 00:30:17,180
right into a current one. So I can see Blair

363
00:30:17,180 –> 00:30:21,210
here is currently viewing the page, and we’ve seen that

364
00:30:21,210 –> 00:30:24,970
current journey. If you think about a movie, maybe even a DVD,

365
00:30:24,970 –> 00:30:28,970
there’s chapters selections, those are the meaningful moments in that

366
00:30:28,970 –> 00:30:32,260
movie. And that’s what we’re seeing here in that unique journey, that

367
00:30:32,310 –> 00:30:36,420
as Blair is navigating through that page, there’s certain points

368
00:30:36,420 –> 00:30:39,350
that we should be aware of and maybe even act on.

369
00:30:41,290 –> 00:30:44,350
On the bottom right we can see scores and this is where

370
00:30:44,350 –> 00:30:48,170
we start using machine learning to drive that engagement. So

371
00:30:48,260 –> 00:30:51,830
when we say predictive engagement with chatbots, what we mean is

372
00:30:51,830 –> 00:30:56,170
driving these chatbots for a business outcome. Using machine learning

373
00:30:56,170 –> 00:31:00,700
to understand how we can increase insurance quotes being driven

374
00:31:00,700 –> 00:31:03,240
on the website itself. And there’s a lot of ways

375
00:31:03,240 –> 00:31:05,880
we can do that, when we tell the system if

376
00:31:05,880 –> 00:31:09,540
someone is navigating to insurance and they navigate away, go

377
00:31:09,540 –> 00:31:13,830
ahead and extend a chatbot within that current journey. Or in

378
00:31:13,830 –> 00:31:16,420
this case, if we look at them using another device,

379
00:31:16,640 –> 00:31:20,150
maybe they’re on a mobile phone? Use something else, reach

380
00:31:20,150 –> 00:31:24,760
out for an SMS or a WhatsApp message, having the flexibility

381
00:31:24,760 –> 00:31:28,020
to change channels gives us that ability to also decide

382
00:31:28,300 –> 00:31:31,930
how we reach out on the device they’re using. So

383
00:31:31,930 –> 00:31:35,080
we’ve realized something here, we’ve got Sarah Marshall on this

384
00:31:35,080 –> 00:31:39,960
page, like I’m mentioning, searching around seeing nothing on car insurance,

385
00:31:40,260 –> 00:31:42,340
I’ve gone to the search box to look for it,

386
00:31:42,630 –> 00:31:45,900
nothing pops up, so I need to engage. And the

387
00:31:45,900 –> 00:31:48,830
best way to do that is with a chatbot. So

388
00:31:48,830 –> 00:31:50,190
we could do it a few ways, we could have

389
00:31:50,190 –> 00:31:52,640
that predictive chatbot come out, I could start a web

390
00:31:52,640 –> 00:31:56,030
chat, and we could reach out to that chatbot first.

391
00:31:56,490 –> 00:32:00,010
It’s important that the chatbot understands what’s going on in

392
00:32:00,010 –> 00:32:03,550
my current journey here. So let’s zoom in and take

393
00:32:03,550 –> 00:32:08,150
a look at that interaction. We’re greeted by K, our virtual assistant

394
00:32:08,150 –> 00:32:11,190
and she sees that I’m interested in auto insurance, by

395
00:32:11,190 –> 00:32:14,630
using that machine learning information that was brought into the chatbot

396
00:32:15,240 –> 00:32:20,270
right away. And that can just be conversational. So I’ve purchased a new

397
00:32:20,270 –> 00:32:28,360
car, any deals for that? Chatbot comes back with some

398
00:32:28,360 –> 00:32:31,230
frequently asked questions and one of those might be a

399
00:32:31,230 –> 00:32:33,500
new car pledge. So what type of deals you can

400
00:32:33,500 –> 00:32:36,490
get when you’ve just gotten a new car. And this

401
00:32:36,490 –> 00:32:39,270
is what we see chatbots use quite a bit and that we

402
00:32:39,270 –> 00:32:42,410
talk about this web design drives engagement through the chat

403
00:32:42,790 –> 00:32:45,770
and we use it for frequently asked questions, but it’s

404
00:32:45,770 –> 00:32:48,870
about taking that further. So what if I have questions

405
00:32:48,870 –> 00:32:53,130
about this specific new car pledge like can I add

406
00:32:53,130 –> 00:33:01,690
a family member? Something that we want chatbots to keep in

407
00:33:01,690 –> 00:33:05,720
mind is that context is always important. If I follow

408
00:33:05,720 –> 00:33:09,230
up with questions, those questions should be attached to the

409
00:33:09,230 –> 00:33:12,850
first intent or what I was originally asking when I

410
00:33:12,850 –> 00:33:16,870
started that interaction. So I’m following up about that new

411
00:33:16,870 –> 00:33:21,010
car pledge, I’m asking about adding family members, and what

412
00:33:21,010 –> 00:33:25,190
about taking that further. Well, frequently we just see chatbots being

413
00:33:25,190 –> 00:33:28,530
used around the idea of frequently asked questions, but what

414
00:33:28,530 –> 00:33:32,250
about automation as well? So let’s ask this chatbot if you can prepare

415
00:33:33,190 –> 00:33:38,080
a quote for me? This is something that we’ve seen

416
00:33:38,080 –> 00:33:41,680
more and more, the idea of process automation within chats.

417
00:33:42,110 –> 00:33:45,950
Agents are, they are evaluable, and if we’re taking up

418
00:33:45,950 –> 00:33:48,810
their time for a repeat task, couldn’t we be using

419
00:33:48,810 –> 00:33:53,220
that on more high impactful moments, customer frustrations, really difficult

420
00:33:53,220 –> 00:33:57,230
questions? What we can do here is drive that initial

421
00:33:57,230 –> 00:34:01,810
interaction like preparing the quote, all from within this digital

422
00:34:01,810 –> 00:34:05,970
channel automated. And that allows me to enter in information

423
00:34:05,970 –> 00:34:09,400
that could be sensitive without sending it over to anyone

424
00:34:09,400 –> 00:34:12,580
else or it allows me to get that quote right

425
00:34:12,580 –> 00:34:14,780
away at a time that we might not even have

426
00:34:14,780 –> 00:34:17,810
our contacts that are open at that moment. So as

427
00:34:17,810 –> 00:34:20,790
I’m driving this engagement, I’m going through that beautiful new

428
00:34:20,790 –> 00:34:23,210
car that I just bought, I can get that quote

429
00:34:23,290 –> 00:34:25,730
right within the chat and I’ll just go ahead and

430
00:34:25,730 –> 00:34:30,200
view it in the window. Now, that’s great, we’ve hit

431
00:34:30,200 –> 00:34:33,030
on a few of those main components as being predicatively

432
00:34:33,030 –> 00:34:37,830
engaged, being context aware on this customer’s current journey, so

433
00:34:37,830 –> 00:34:40,510
if we come back to that Sarah Marshall engagement within

434
00:34:40,510 –> 00:34:43,810
the Altocloud side, we can see that full web journey

435
00:34:43,810 –> 00:34:48,510
that’s happened here, that drove that chatbot because we know that this

436
00:34:48,510 –> 00:34:51,250
person was interested in quotes and they were searching for

437
00:34:51,250 –> 00:34:55,460
car insurance. But let’s go ahead and end here. I’ve

438
00:34:55,460 –> 00:34:58,410
gotten the information I need for now and maybe time

439
00:34:58,410 –> 00:35:00,469
passed, I’m not quite sure if I’m ready to sign

440
00:35:00,469 –> 00:35:03,410
up or maybe I have signed up, things could happen.

441
00:35:04,330 –> 00:35:06,870
It’s important that we don’t stay to that one channel,

442
00:35:06,870 –> 00:35:09,730
and that’s something we mentioned earlier but how do we

443
00:35:09,730 –> 00:35:12,650
do that? On the case of a car insurance quote,

444
00:35:12,950 –> 00:35:15,730
we might go ahead and send someone an email, so

445
00:35:15,730 –> 00:35:17,610
I’m going to mirror my cellphone here for just a

446
00:35:17,610 –> 00:35:20,920
moment. I want to show that when we have these interactions,

447
00:35:20,969 –> 00:35:24,610
automation isn’t just within the chat itself, it’s about following

448
00:35:24,610 –> 00:35:27,100
up. In this case we could have reached out to

449
00:35:27,100 –> 00:35:30,980
Sarah with an email and showing that car insurance quote,

450
00:35:31,360 –> 00:35:34,190
so that I can go ahead and process the payment right

451
00:35:34,190 –> 00:35:36,980
with my phone in front of me. And as we talk

452
00:35:37,170 –> 00:35:40,250
about Bots, it’s important that we actually have the ability

453
00:35:40,250 –> 00:35:44,150
to extrapolate or take certain pieces of that Bot and

454
00:35:44,150 –> 00:35:46,880
use it in a siloed manner. What we have here

455
00:35:46,880 –> 00:35:50,280
is what we call a micro application. So there’s certain

456
00:35:50,280 –> 00:35:52,860
pieces of the Bot like preparing the quote that you saw

457
00:35:52,860 –> 00:35:55,880
here, that I might want to just make accessible to

458
00:35:55,880 –> 00:35:58,700
a customer in a more siloed fashion. In this case,

459
00:35:58,700 –> 00:36:00,760
I could go ahead and pay that quote if I

460
00:36:00,760 –> 00:36:05,860
wanted to, by just entering in some information and going

461
00:36:05,860 –> 00:36:09,340
through that process back and forth. What if I’ve still

462
00:36:09,340 –> 00:36:13,920
got some questions? Let’s go back to that email. We

463
00:36:13,920 –> 00:36:16,810
talk about Bots and we talk about channels, or what

464
00:36:16,810 –> 00:36:18,420
if I just want to call in and ask these

465
00:36:18,450 –> 00:36:23,080
questions? Voicebots should be just as accessible as we see

466
00:36:23,080 –> 00:36:25,760
within a chatbot. And let’s take a look at that

467
00:36:25,760 –> 00:36:28,500
here. I’m going to go ahead and make a phone

468
00:36:28,500 –> 00:36:31,460
call and we’ll ask some follow up questions about that

469
00:36:31,460 –> 00:36:34,160
car insurance quote just to see how I can get

470
00:36:34,160 –> 00:36:37,989
help? Potentially even make changes and then transfer to an

471
00:36:37,989 –> 00:36:50,540
agent if I’d like. Hi Sarah. Thank you for calling

472
00:36:50,540 –> 00:36:54,840
G Insurance. How can I help? Hey there. How does

473
00:36:54,840 –> 00:37:00,840
that deductible work? Okay. The following has been found for

474
00:37:00,840 –> 00:37:04,520
your query. A car insurance deductible is the amount of

475
00:37:04,520 –> 00:37:08,350
money you have to pay toward repair your insurance covers the rest.

476
00:37:08,750 –> 00:37:11,050
You have to pay a deductible no matter who is

477
00:37:11,050 –> 00:37:14,980
at fault. In most cases, the higher the deductible, the

478
00:37:14,980 –> 00:37:18,130
lower the premium. Is there anything else I can help

479
00:37:18,130 –> 00:37:24,580
you with? Yeah. How can I help? If I make

480
00:37:24,580 –> 00:37:27,060
a change with my deductible, how would that affect my

481
00:37:27,060 –> 00:37:33,219
monthly payments? Hey, I told you I am a smart

482
00:37:33,430 –> 00:37:37,730
assistant. In most cases, the higher the deductible, the lower

483
00:37:37,730 –> 00:37:41,660
the monthly payments. Your policy allows deductible to be set

484
00:37:41,660 –> 00:37:47,150
to 1, 000, 1, 500 or 2, 000 dollars. Currently your deductible is

485
00:37:47,150 –> 00:37:50,370
1, 000. Is there anything else I can help you

486
00:37:50,370 –> 00:37:57,880
with? Yes. How can I help? Can we go ahead and increase my deductible by

487
00:37:57,880 –> 00:38:05,030
500 dollars? Okay, all set. Your monthly payment will go

488
00:38:05,030 –> 00:38:08,530
down by 50 US dollars. The link to the updated

489
00:38:08,530 –> 00:38:12,400
policy will be sent via SMS. Is there anything else

490
00:38:12,400 –> 00:38:15,140
I can help you with? Yeah, I’d like to speak to

491
00:38:15,550 –> 00:38:20,550
an advisor. You would like to speak to an advisor,

492
00:38:20,630 –> 00:38:26,010
is that correct? Yes. One moment please while I put

493
00:38:26,010 –> 00:38:29,980
you through to an advisor. So let’s take that off

494
00:38:29,980 –> 00:38:31,770
speaker for a moment just to talk about some of

495
00:38:31,770 –> 00:38:35,060
the things that we did here within that Bot interaction. We

496
00:38:35,060 –> 00:38:39,260
went ahead, asked some questions, received responses that were actually

497
00:38:39,260 –> 00:38:43,930
tailored to our current deductible in this situation and we

498
00:38:43,930 –> 00:38:47,770
even went ahead and made a deductible change. Now, we’re going to

499
00:38:47,770 –> 00:38:50,170
reach out to an agent. So as I switch over

500
00:38:50,170 –> 00:38:52,190
my screen here for a moment, we’ll see that we have

501
00:38:52,190 –> 00:38:55,390
our workspace web view, the agent view. And when I

502
00:38:55,390 –> 00:38:59,600
go ready, I’m going to receive that interaction. One of

503
00:38:59,600 –> 00:39:01,690
the things that we don’t hit on all the time that

504
00:39:01,690 –> 00:39:04,810
are so great about Bots is the ability to do

505
00:39:04,890 –> 00:39:08,900
interactions steering. Well, what is this interaction about? Who’s the

506
00:39:08,900 –> 00:39:11,460
best person to help you? How do we get the

507
00:39:11,460 –> 00:39:14,910
information we need to make those decisions? That conversation I

508
00:39:14,910 –> 00:39:18,870
just had, that was the same as navigating an old school

509
00:39:18,730 –> 00:39:22,910
IVR, where I might be entering in some buttons, pressing all of

510
00:39:22,910 –> 00:39:27,440
these boring menu options, instead I’m having a conversation and

511
00:39:27,440 –> 00:39:31,050
that routes me to the right agent. Even better, that

512
00:39:31,050 –> 00:39:35,000
agent knows exactly what this interaction’s about, I even have the

513
00:39:35,000 –> 00:39:39,410
ability to not only receive that customer’s information, but quickly drive into

514
00:39:39,410 –> 00:39:42,840
those interactions. So if you remember that chat we had

515
00:39:42,840 –> 00:39:45,590
earlier, well, I can review all of that right on

516
00:39:45,590 –> 00:39:49,070
my screen from an agent perspective, even review the voicebot

517
00:39:49,280 –> 00:39:51,910
interaction that we just had a moment ago. All of

518
00:39:51,910 –> 00:39:55,480
this should be accessible to us because again, we need

519
00:39:55,480 –> 00:39:57,280
to make sure that when we transfer to the agent,

520
00:39:57,700 –> 00:40:01,290
they have full awareness on those past interactions. So not

521
00:40:01,290 –> 00:40:04,410
just what happened yesterday, but what happened across every single

522
00:40:04,410 –> 00:40:09,030
channel. And this is important about Bots because we’re quickly realizing

523
00:40:09,030 –> 00:40:11,830
that customers respond to companies as if they’re a single

524
00:40:11,830 –> 00:40:15,270
person. They might say I spoke to you yesterday or

525
00:40:15,660 –> 00:40:18,600
I chatted with you earlier about this insurance policy, don’t you

526
00:40:18,840 –> 00:40:22,170
about that? It’s important that your agents have full awareness

527
00:40:22,170 –> 00:40:24,440
of all the power that we’re giving bots as well.

528
00:40:26,870 –> 00:40:29,700
So to come back to that phone again, we’re on

529
00:40:29,700 –> 00:40:32,500
that phone call conversation, the agents got all the information

530
00:40:32,500 –> 00:40:35,080
they need and I received that text message like we

531
00:40:35,080 –> 00:40:38,850
talked about earlier that automatically showed me that new policy

532
00:40:39,060 –> 00:40:41,190
and I could reference that while I’m on the phone

533
00:40:41,190 –> 00:40:45,500
call with an agent. So let’s go ahead and end

534
00:40:45,500 –> 00:40:49,630
this interaction here and we’ll tie this up and I want to show

535
00:40:49,630 –> 00:40:52,739
one more Bot use case as we talk about all

536
00:40:52,739 –> 00:40:54,830
the different ways that we could be engaging and all of

537
00:40:54,830 –> 00:40:58,410
the ways that customers are expecting from us. Another big

538
00:40:58,410 –> 00:41:02,219
one is social media or different channels in general. Something

539
00:41:02,219 –> 00:41:05,130
to think about is how our customers are engaging with

540
00:41:05,130 –> 00:41:08,960
us? And Facebook messenger and WhatsApp and a business chat,

541
00:41:08,960 –> 00:41:11,480
these are all new ways that we need to be

542
00:41:11,480 –> 00:41:15,239
not only available but using that same Bot technology or

543
00:41:15,239 –> 00:41:19,540
that same orchestration there. Now, one of our key principles

544
00:41:19,800 –> 00:41:24,330
was being and Facebook has some great thoughts as well on the

545
00:41:24,410 –> 00:41:27,550
natural language understanding side, what you may have heard as

546
00:41:27,610 –> 00:41:31,290
Wit or Wit. ai. It’s important that you be able

547
00:41:31,290 –> 00:41:33,960
to use those as well. So if I were to

548
00:41:33,960 –> 00:41:36,989
reach out to G Insurance on Facebook, they might be

549
00:41:36,989 –> 00:41:40,410
using an entirely different NLU on the backend. It might

550
00:41:40,410 –> 00:41:42,510
have a bit more small talk, it might be a

551
00:41:42,510 –> 00:41:45,160
little bit better at just going back and forth with

552
00:41:45,160 –> 00:41:49,210
customers in a casual conversation, it’s still important that it

553
00:41:49,210 –> 00:41:51,980
has the ability to access all the data on me

554
00:41:52,360 –> 00:41:55,070
just as we saw within the chat, just as we saw

555
00:41:55,070 –> 00:41:58,730
within that voice call. So in this situation, let’s say

556
00:41:58,730 –> 00:42:01,500
sometime goes by and I reach out to G Insurance

557
00:42:01,500 –> 00:42:03,510
on Facebook because I’m on the side of the road

558
00:42:03,969 –> 00:42:06,520
and maybe I forgot a crack in my windshield, so

559
00:42:06,520 –> 00:42:10,700
I need help. Let me go ahead and engage here,

560
00:42:12,300 –> 00:42:15,160
one thing I always like to do is prove that

561
00:42:15,160 –> 00:42:17,810
it’s a live demo. Every once in a while you

562
00:42:17,810 –> 00:42:20,010
might see some fun that pops up but this is how

563
00:42:20,110 –> 00:42:22,880
we know that it’s live. Let me get a little

564
00:42:23,010 –> 00:42:30,760
bit more aggressive with it. Well, Facebook might be sleeping

565
00:42:30,760 –> 00:42:33,320
today, but what we wanted to show here was when

566
00:42:33,320 –> 00:42:37,090
we engage on that G Insurance Bot that’s within Facebook,

567
00:42:37,350 –> 00:42:40,170
we can actually go ahead and set up information like

568
00:42:40,170 –> 00:42:43,870
getting your windshield fixed. So going through that automated journey

569
00:42:43,870 –> 00:42:47,540
there and having that available to you right within another

570
00:42:47,540 –> 00:42:50,880
channel. So situations that, maybe I’m on the road, maybe

571
00:42:50,880 –> 00:42:53,070
I’m at work and I need help, I might not

572
00:42:53,070 –> 00:42:55,170
be able to reach out on that web chat, I

573
00:42:55,170 –> 00:42:57,870
might not be able to call in right away, I’m actually going to go ahead and try

574
00:42:58,790 –> 00:43:02,520
to do something like Facebook. So apology is there, that’s

575
00:43:02,750 –> 00:43:06,520
always the nature with live demos but from here, we

576
00:43:06,520 –> 00:43:09,590
can start to open this up to some Q&A if anyone has

577
00:43:09,590 –> 00:43:15,550
any and let’s see. Actually, before we open that to Q&A, I

578
00:43:15,550 –> 00:43:19,210
think I’m going to give it one more go. One of the things I’ve realized

579
00:43:19,210 –> 00:43:21,870
in the past is if I have my agent available,

580
00:43:21,870 –> 00:43:25,730
it might be ringing me so that’s on me. Let’s

581
00:43:25,730 –> 00:43:32,270
try one more time and we’ll bring up that screen

582
00:43:32,270 –> 00:43:49,719
share. Okay. Well, we tried. So at this point, we can

583
00:43:49,719 –> 00:43:52,510
go back, let’s open up to some Q& A guys, my apologies there.

584
00:43:52,580 –> 00:43:54,460
We can always follow up with another demo and show

585
00:43:54,460 –> 00:43:57,150
that Facebook piece. I’d love to talk further if that’s

586
00:43:57,150 –> 00:44:04,120
something you’re interested in. No problem. Thanks Joe for the

587
00:44:04,120 –> 00:44:07,219
excellent demo that you did provide. And like he said,

588
00:44:07,670 –> 00:44:09,080
go and reach out to us if you’d like to

589
00:44:09,080 –> 00:44:13,770
see that more in depth by either reaching out to your executive

590
00:44:13,770 –> 00:44:18,420
or having an SDR follow up. I think we’d like to

591
00:44:18,420 –> 00:44:22,680
show you that in a personal demo environment. Like he

592
00:44:22,680 –> 00:44:24,350
said, we’re actually go ahead and open up to the

593
00:44:25,310 –> 00:44:28,270
Q& A. With that being said, to participate in today’s

594
00:44:28,270 –> 00:44:30,400
Q& A, go ahead and throw those questions into the

595
00:44:30,400 –> 00:44:32,730
Q& A window in the middle and we’ll answer as

596
00:44:32,730 –> 00:44:34,330
many as we can with the time that we have

597
00:44:34,330 –> 00:44:40,940
remaining. First up, I’m going to ask a question to

598
00:44:41,280 –> 00:44:48,270
Torsten. Torsten, our friend Kelly wants to know, do the

599
00:44:48,270 –> 00:44:53,550
Bosch Bots support things like voice chat, both or something

600
00:44:53,550 –> 00:44:55,680
else? Could you expand a little bit on what your

601
00:44:55,680 –> 00:45:01,590
Bots currently support? Yeah. So we started with for web chat Bots

602
00:45:02,290 –> 00:45:05,620
because of the huge demand and for us to other

603
00:45:05,620 –> 00:45:11,590
channels. So Bots still text based and we’re already investigating and did some

604
00:45:11,590 –> 00:45:16,340
proof of concept with a voice part as well. Really identifying also the challenges when it comes to Bot vendors and this kind of

605
00:45:21,170 –> 00:45:26,239
boring topics which could be very stressful in real production.

606
00:45:26,239 –> 00:45:32,810
So the production use cases are really text based Bots

607
00:45:33,000 –> 00:45:37,080
but voices also a huge demand which is coming up

608
00:45:37,080 –> 00:45:43,030
more and more and as Joe showed. There’s also a reason for

609
00:45:43,030 –> 00:46:01,260
this so kind of moving away from traditional TGMF and asked by the Dialogs to Bot conversation rooms. Why should I have this different experience in kind of established channel like why should I have a poor experience when

610
00:46:01,260 –> 00:46:03,790
I just came from a webpage where I had this

611
00:46:04,020 –> 00:46:08,550
Bot conversation or really a conversation with a company already?

612
00:46:09,920 –> 00:46:15,110
Yeah, absolutely. I think that’s such a great point Torsten. People become not

613
00:46:15,110 –> 00:46:18,160
reliant but they expect that experience. Especially when you give

614
00:46:18,160 –> 00:46:21,150
them something that’s great, that should be consistent across those

615
00:46:21,150 –> 00:46:28,960
other channels as well. Awesome. Well, hopefully that answers your

616
00:46:28,960 –> 00:46:30,930
question Kelly. If not you can go ahead and throw

617
00:46:30,930 –> 00:46:34,060
a follow up into the Q& A window and I’m going to go ahead

618
00:46:34,060 –> 00:46:36,920
and move onto our friend Mike here, he would like

619
00:46:36,920 –> 00:46:39,510
to know, can you speak to the importance of the

620
00:46:39,550 –> 00:46:43,980
NLU in Bots to handle the 35 different languages you

621
00:46:43,980 –> 00:46:48,580
have to support and was this a challenge to their

622
00:46:48,580 –> 00:47:06,570
fine tuning in your Bots? Yeah. So the beneficial part of our business is we don’t need to cover all these languages in one customer service project. So we have customers out of the different regions, the answer is we were not face this challenge, but due to the amount

623
00:47:06,739 –> 00:47:09,920
of countries we’re supporting, we definitely had this as a

624
00:47:10,510 –> 00:47:17,130
huge topic when evaluating the Bots and also, the variety of Bots let’s

625
00:47:18,920 –> 00:47:22,700
just call it Bot integrations, we’re already offering kind of,

626
00:47:22,900 –> 00:47:29,780
protects us from, customer asks for a country or language

627
00:47:29,780 –> 00:47:32,310
support and the Bot we are having in our portfolio

628
00:47:32,310 –> 00:47:37,360
doesn’t support it. So the beneficial part is that we are in

629
00:47:37,750 –> 00:47:42,040
the position now to really pick the Bots based on the customer

630
00:47:42,550 –> 00:47:46,630
requirements or what I outlined before, we are even pushed

631
00:47:46,730 –> 00:47:50,180
to integrate a different Bot maybe from a customer coming from a

632
00:47:50,480 –> 00:47:54,230
region where they already did this evaluation on their own

633
00:47:54,230 –> 00:48:00,430
and asked us, okay, please use this, choose a Bot from us to

634
00:48:00,430 –> 00:48:07,260
integrate it or to deliver this in your customer service. You’re either offering to us so this kind of independence from the Bot vendors also

635
00:48:07,260 –> 00:48:14,180
kind of give us the freedom we need to pick

636
00:48:15,010 –> 00:48:27,180
the best matching Bot for the specific customer requirements without having to redesign our full platform or even changing a lot of things in the

637
00:48:27,180 –> 00:48:36,960
business logic so really the capabilities and abilities to, in worse case

638
00:48:36,960 –> 00:48:41,530
you mix different Bots. So depending on some, as I said,

639
00:48:41,530 –> 00:48:46,330
routing decisions, that’s kind of unique for us and I

640
00:48:46,330 –> 00:48:49,290
said it’s kind of a very valuable feature in the

641
00:48:49,460 –> 00:48:54,489
platform we have now. There were also questions towards compliance and so and so,

642
00:48:56,280 –> 00:49:02,380
what I mentioned at the beginning, because we’re in different countries,

643
00:49:02,380 –> 00:49:06,680
we are also offering internal services so we have a variety

644
00:49:06,680 –> 00:49:12,350
of different legal and also on top, customer requirements, which

645
00:49:12,350 –> 00:49:20,150
can become legal requirements. So we can currently fully deliver on

646
00:49:20,150 –> 00:49:26,800
Premise, Bot experience without any other sub processors Bosch. So it’s fully

647
00:49:26,800 –> 00:49:31,820
on Premise, so that’s possible why we are able also to

648
00:49:31,820 –> 00:49:36,860
integrate a new Cloud Bot provider or AI provider. At

649
00:49:36,860 –> 00:49:40,440
the same time without, as I said, tearing down the full

650
00:49:40,440 –> 00:49:45,060
platform and implementing something from scratch. Yeah. And one thing

651
00:49:45,060 –> 00:49:48,610
to know is, I think what’s super powerful is the ability to

652
00:49:48,610 –> 00:49:51,020
run multiple Bots in parallel. So you could have a

653
00:49:51,020 –> 00:49:54,440
situation where the Bot is interacting with someone and maybe

654
00:49:54,440 –> 00:49:57,540
that Bot only understands one language, but you could layer

655
00:49:57,540 –> 00:50:00,420
another Bot that it’s actually just there for translation services

656
00:50:00,739 –> 00:50:03,860
and we’ve seen that used even internally, we have lead

657
00:50:03,860 –> 00:50:07,530
development reps that might be based in South America, maybe

658
00:50:07,530 –> 00:50:09,969
they only speak Spanish, not only speak English and a

659
00:50:09,969 –> 00:50:13,570
Portuguese chat comes in. Well, that can throw a wrench

660
00:50:13,570 –> 00:50:16,230
in the system there, so the ability to have a

661
00:50:16,230 –> 00:50:20,530
Bot that overlays, has ability to give the natural language

662
00:50:20,530 –> 00:50:23,080
that agent and then give the respective language to that

663
00:50:23,080 –> 00:50:31,580
customer or person reaching out, that’s really important as well. Awesome. Thanks

664
00:50:31,580 –> 00:50:35,739
guys. I’m going to direct this next question to Joe, our

665
00:50:35,739 –> 00:50:40,760
friend Clayton wants to know, does Kate require intelligent automation

666
00:50:40,760 –> 00:50:44,330
to work or could you go into a little bit more detail about what

667
00:50:44,330 –> 00:50:48,380
Kate does require? Absolutely. And Clayton, that’s a great question.

668
00:50:48,380 –> 00:50:51,469
So when we think about Kate, that is our virtual assistant

669
00:50:51,469 –> 00:50:54,010
here and Kate could manifest in a lot of ways

670
00:50:54,010 –> 00:50:56,880
and might use many different Bots in the backend depending on

671
00:50:56,880 –> 00:51:02,739
your business. Intelligent automation is one of the tools that we use within

672
00:51:02,820 –> 00:51:05,460
Pure Engage which is our platform at Genesys. So we’ve

673
00:51:05,460 –> 00:51:09,320
got three platforms here. That you might be a PureCloud,

674
00:51:10,320 –> 00:51:14,040
PureConnect or PureEngage customer or someone looking into those. They’ve

675
00:51:14,040 –> 00:51:16,989
all got unique capabilities in terms of helping your business

676
00:51:16,989 –> 00:51:21,500
from a customer experience stand point. Intelligent automation is one

677
00:51:21,500 –> 00:51:25,750
of those tools within PureEngage, if you’re using PureCloud, you

678
00:51:25,750 –> 00:51:29,200
might be using something called Architect. So it’s not required,

679
00:51:29,200 –> 00:51:31,760
it’s simply a piece of the puzzle if it’s on

680
00:51:31,760 –> 00:51:35,040
that platform for you. Basically just means that if you’re

681
00:51:35,040 –> 00:51:38,190
doing that type of work in that platform, that’s one of the

682
00:51:38,190 –> 00:51:41,239
tools we use, but I don’t want you to think that

683
00:51:41,300 –> 00:51:43,420
you’re only tied to that one. So depending on the

684
00:51:43,420 –> 00:51:46,300
platform you’re using, you might have some different choices there.

685
00:51:50,250 –> 00:51:54,180
Thanks Clayton for your question, hopefully that answers that. I’m

686
00:51:54,180 –> 00:51:56,500
going to go ahead and move on to a question

687
00:51:56,500 –> 00:52:03,460
for Torsten here. Okay. The question is, is it possible

688
00:52:03,460 –> 00:52:08,750
to integrate information from Bosch devices in chat? In addition

689
00:52:08,750 –> 00:52:12,590
to that, what roles at Bosch Service Solutions have the

690
00:52:12,600 –> 00:52:16,239
highest workload when it comes to maintenance and training these

691
00:52:16,239 –> 00:52:18,100
Bots? Could you expand a little bit more on that?

692
00:52:19,780 –> 00:52:22,560
Yep. So for the first question, I think you’ve watched the

693
00:52:23,110 –> 00:52:28,660
IoT commercials, so I think currently not. Not yet. So

694
00:52:29,100 –> 00:52:35,480
the Bosch Global webpage is one of our customers for example so the extension of

695
00:52:35,960 –> 00:52:41,300
this or integrating IoT capable devices, they have mobile devices and so

696
00:52:41,300 –> 00:52:46,300
on. That’s definitely something we see as a demand, we

697
00:52:46,300 –> 00:52:57,750
will face soon because also you asked. So the customer demand for these integrations, that’s predictable already.

698
00:52:59,980 –> 00:53:02,770
We don’t see any, well, we don’t have any fear

699
00:53:02,770 –> 00:53:06,850
for this, so I think we are fine with also kind of

700
00:53:06,850 –> 00:53:10,760
this additional information. So also they are, I would question myself, okay, is this kind

701
00:53:16,719 –> 00:53:22,320
of technical information or whatever part for the routing, already

702
00:53:22,320 –> 00:54:01,920
for the Bot or both? To really find out what’s the best handling part for this or even at some RPA mechanisms for this and how to reintegrate them. But I don’t see the, for us it would be an additional channel with additional content. So nothing really, where we, as I said before, where we need to start from scratch. It has some major implications, but more to load and other things. So it’s additional contents which you need to handle somehow. Maybe you have an IoT Bot, which we would then add as

703
00:54:02,020 –> 00:54:07,360
well. So these possibilities, we think we’re on the safe

704
00:54:07,360 –> 00:54:13,010
side to add this capabilities. And the other question was

705
00:54:13,210 –> 00:54:18,000
the, where’s the highest workloads? So this was really an

706
00:54:18,070 –> 00:54:22,150
interesting discussion also, who’s really handling the maintenance of the

707
00:54:22,150 –> 00:54:30,360
Bots? This is more outside of IT so we’re kind of identifying

708
00:54:30,360 –> 00:54:42,680
is the Bot answering well or do I have a

709
00:54:42,680 –> 00:54:44,680
lot of unanswered questions? That’s more driven by the business. So the input is driven by the business, assisted because most of the time

710
00:54:44,680 –> 00:54:50,989
we started with an existing web chat, so we had some content we could already kind of identify that’s the probability, which could be

711
00:54:54,469 –> 00:55:18,300
high for this topics and then really business and team leaders are working on identifying what’s maybe missing in the Bot, what’s going wrong? And then a collaboration with IT and other stakeholders happens. Awesome. So unfortunately we only have

712
00:55:18,300 –> 00:55:21,160
time for about one more question before I end today’s

713
00:55:21,160 –> 00:55:24,140
session. However, don’t fret, if we didn’t answer your question

714
00:55:24,140 –> 00:55:26,570
here in the live session, we will follow up with

715
00:55:26,570 –> 00:55:29,630
you via email within the next few days. So this last question is going to go

716
00:55:31,120 –> 00:55:33,270
to you Torsten and if there’s anything you’d like to

717
00:55:33,270 –> 00:55:39,300
add Joe, feel free. So Torsten, our friend Annie wants

718
00:55:39,300 –> 00:55:42,500
to know, could you talk a little bit more about

719
00:55:42,500 –> 00:55:45,350
how many intents that you’ve had to create or how

720
00:55:45,350 –> 00:55:48,620
many do you expect to add? And talk about maybe

721
00:55:48,830 –> 00:55:56,280
the staff that you’ve dedicated to enhance these Bots? Yeah.

722
00:55:56,910 –> 00:56:01,350
As I said before, because of all the businesses it’s

723
00:56:01,350 –> 00:56:09,560
really segmented. So we have multiple projects with multiple Bots. So there’s no general answer because there’s no Bosch

724
00:56:09,560 –> 00:56:12,660
Service Solutions answer. Every answer would be kind of project

725
00:56:12,660 –> 00:56:21,260
or a customer specific for us. So maybe Joe has some more or a kind of a number or an

726
00:56:21,260 –> 00:56:25,660
answer which applies more to you? No, that’s a great

727
00:56:25,660 –> 00:56:29,130
question. I think, as Torsten said, the intense you have

728
00:56:29,130 –> 00:56:31,090
and just for everyone in the audience here, the intense

729
00:56:31,090 –> 00:56:35,670
would be, what are we trying to accomplish from that

730
00:56:35,670 –> 00:56:39,910
customer interactions? So if I said something like, I need to

731
00:56:40,660 –> 00:56:44,580
change my deductible, the intent would be changing deductible there.

732
00:56:45,510 –> 00:56:48,450
So for a business, that could be entirely different, you

733
00:56:48,450 –> 00:56:51,020
may find that you need a lot of intense, you

734
00:56:51,020 –> 00:56:53,190
may find you only have three or four, depending on

735
00:56:53,190 –> 00:56:56,640
your business. An important piece to go with that is

736
00:56:56,640 –> 00:56:59,610
training the Bot to understand how’d you get to that

737
00:56:59,610 –> 00:57:03,180
intent? So I’m like, say I need to change my deductible, I

738
00:57:03,180 –> 00:57:06,190
might also say, hey, I don’t want to pay that

739
00:57:06,190 –> 00:57:07,960
much when I get into an accident, and that could

740
00:57:07,960 –> 00:57:10,760
mean the same thing. And that’s where I think, the

741
00:57:10,760 –> 00:57:14,540
importance of adding a lot or adding more is really

742
00:57:14,540 –> 00:57:19,970
going to show within the effect of of your Bot. Yep.

743
00:57:22,550 –> 00:57:25,980
I think at the end it’s really also not that critical. As long as you

744
00:57:25,980 –> 00:57:32,840
allow this proper transition to the agent, even if you start really small and you

745
00:57:32,840 –> 00:57:37,320
don’t cover the full range, you still have at least

746
00:57:37,320 –> 00:57:40,660
partly understood what’s happening and you have a coverage of

747
00:57:42,630 –> 00:57:46,450
the interactions and the ones you’re not covering, the agent

748
00:57:46,450 –> 00:57:50,040
can also give some feedback towards the Bot training or

749
00:57:50,040 –> 00:57:55,770
whatever. But there’s also something we did that really used the

750
00:57:55,770 –> 00:58:00,820
agents back to, in cycles so the team leaders to get this input

751
00:58:00,820 –> 00:58:04,090
as well. I love that Torsten. I think, we don’t

752
00:58:04,090 –> 00:58:06,980
want someone to have a guillotine moment where they’re chopped

753
00:58:06,980 –> 00:58:09,450
and they’re done, the interaction’s over, there’s no way to

754
00:58:09,450 –> 00:58:16,920
get more information so that’s absolutely crucial. Well, thanks guys

755
00:58:17,100 –> 00:58:21,070
for providing your awesome answers to excellent questions from our

756
00:58:21,070 –> 00:58:25,500
live audience. Unfortunately that is all the time that we

757
00:58:25,500 –> 00:58:28,590
have today for Q& A, but like I said before,

758
00:58:28,590 –> 00:58:31,790
go ahead and throw those questions into the Q& A window before we end

759
00:58:32,040 –> 00:58:34,790
today’s session and we’ll be sure to follow up with

760
00:58:34,790 –> 00:58:38,690
you via email or feel free to email us specifically

761
00:58:38,690 –> 00:58:42,880
at [email protected] genesys. com, no caps or spaces, with the

762
00:58:42,880 –> 00:58:45,600
title of the webcast in the subject line and we’ll

763
00:58:45,600 –> 00:58:50,130
be sure to respond promptly. So with all that being

764
00:58:50,130 –> 00:58:53,540
said, to wrap up first, don’t forget to take advantage

765
00:58:53,540 –> 00:58:56,750
of the additional resources within that resource list below the

766
00:58:56,750 –> 00:59:02,170
Q& A window. Clicking, before I end today’s session, we’ll

767
00:59:02,170 –> 00:59:04,990
open up in a new tab and that’ll really expand

768
00:59:04,990 –> 00:59:09,150
on today’s topic, giving you some additional information. And also

769
00:59:09,150 –> 00:59:11,860
the friendly reminder, if you could fill up that survey that’s

770
00:59:11,860 –> 00:59:13,710
going to show up at the end of the webcast,

771
00:59:13,710 –> 00:59:16,900
providing your thoughts on today’s presentation, we’d greatly appreciate that

772
00:59:16,900 –> 00:59:22,060
feedback. So on behalf of Torsten, Joe and the entire

773
00:59:22,060 –> 00:59:25,860
Genesys team, we thank you again for joining today’s live

774
00:59:25,860 –> 00:59:30,080
webcast, five things to consider when building chatbots. Until next

775
00:59:30,080 –> 00:59:31,720
time, have a good one everyone.