1
00:00:00,480 --> 00:00:01,480
Equidox.  By Onix.  

2
00:00:01,600 --> 00:00:04,480
Reach Everyone

3
00:00:05,000 --> 00:00:10,160
[Dan Tuleta] Okay so it's just about two 
o'clock, so I think we're ready to get started.  

4
00:00:11,040 --> 00:00:15,040
For everyone that's still shuffling in, all this 
is being recorded so you won't miss too much in  

5
00:00:15,040 --> 00:00:20,880
the very first minute or two. Thank you all for 
attending. Welcome again for those of you who  

6
00:00:20,880 --> 00:00:25,440
have attended our webinars before. This is the 
next installment of Equidox Webinar Wednesdays.  

7
00:00:27,680 --> 00:00:33,840
I heard a bit of background noise there I’m not 
sure if someone was off of mute... Okay so welcome  

8
00:00:33,840 --> 00:00:38,480
everyone to Equidox Webinar Wednesdays. Talking 
about our newest service, batch processing.  

9
00:00:39,920 --> 00:00:45,200
As always we do appreciate your attendance and 
if you have any questions that we do not answer  

10
00:00:45,200 --> 00:00:48,720
during this presentation, please 
feel free to reach out to us...  

11
00:00:50,560 --> 00:00:59,120
[Is there an issue with the audio? Is that me? 
I'm not sure… There's a whole bunch of .. I’m  

12
00:00:59,120 --> 00:01:05,360
hearing people like entering and leaving…] but 
anyways... Thank you again for joining us today,  

13
00:01:05,360 --> 00:01:08,960
and if you have any questions throughout this 
presentation that we do not cover, please do  

14
00:01:08,960 --> 00:01:14,880
not hesitate to reach out to us either through 
our website at www.Equidox.co. You can call us  

15
00:01:14,880 --> 00:01:22,080
directly at 800-664-9638, or the best way probably 
to reach out initially would be through our email  

16
00:01:22,080 --> 00:01:29,040
which can be found at EquidoxSales@Onixnet.com. 
We're also very active on LinkedIn and social  

17
00:01:29,040 --> 00:01:33,680
media, so if you are if you're on LinkedIn 
please feel free to connect with us. We post a  

18
00:01:33,680 --> 00:01:39,840
lot of articles and information about our product 
services and the accessibility space in general.

19
00:01:42,640 --> 00:01:48,880
So just a couple of quick notes about us before 
we get into the nitty-gritty details of batch  

20
00:01:48,880 --> 00:01:55,440
processing. For those of you who might only know 
us by Equidox, we actually have a parent company  

21
00:01:55,440 --> 00:02:02,320
called Onix Networking and Onix was founded in 
1992 in Cleveland, Ohio. So we're approaching  

22
00:02:02,320 --> 00:02:09,520
nearly 30 years of business in the IT space. We're 
best known for our partnerships with Google and  

23
00:02:09,520 --> 00:02:16,000
AWS (Amazon Web Services). So we are really just 
an all-around cloud consultancy and cloud experts,  

24
00:02:16,560 --> 00:02:21,040
with our mission being to improve organizational 
efficiency through cloud computing solutions.  

25
00:02:21,760 --> 00:02:26,720
So if you if your organization outside of 
accessibility has any interest in having  

26
00:02:26,720 --> 00:02:31,920
conversations about cloud, about cloud software, 
or cloud technology please feel free to reach out  

27
00:02:31,920 --> 00:02:37,600
to us as well. We can definitely direct you to 
the correct people within Onix. Now Equidox is a  

28
00:02:38,560 --> 00:02:43,840
branch of Onix focusing on accessibility. 
So many of you are probably aware of our  

29
00:02:43,840 --> 00:02:50,080
best-in-class PDF remediation software called 
Equidox. We also offer PDF remediation services  

30
00:02:50,080 --> 00:02:54,480
where clients will send us PDF documents 
that they need to have remediated,  

31
00:02:54,480 --> 00:02:59,520
Our team of remediators and validators will 
make those documents accessible, and then we  

32
00:02:59,520 --> 00:03:04,960
will return those documents to our clients, and 
then they can be posted online in an accessible  

33
00:03:04,960 --> 00:03:11,920
format. Obviously mitigating all legal risks to 
comply with accessibility laws and regulations.  

34
00:03:12,800 --> 00:03:19,120
We also offer expert accessibility services which 
can include testing, training, VPAT completion,  

35
00:03:19,120 --> 00:03:23,680
auditing websites, helping you put together 
accessibility plans for your organization. And  

36
00:03:23,680 --> 00:03:28,800
our mission is to ensure that digital information 
reaches everyone through accessibility solutions.

37
00:03:31,600 --> 00:03:36,640
So here are just a few customers who we support 
either through Onix’s cloud business or through  

38
00:03:36,640 --> 00:03:41,920
Equidox, or possibly a combination of both. 
So we work with organizations of all sizes  

39
00:03:41,920 --> 00:03:46,800
in every vertical, so if you have PDF documents 
and you need to make them... and you need to make  

40
00:03:46,800 --> 00:03:52,000
them compliant, the Equidox team can definitely 
help you improve your workflow. So like I said,  

41
00:03:52,000 --> 00:03:57,200
any sort of PDF challenges that you may have we 
are probably the people to talk to. So please  

42
00:03:57,200 --> 00:04:02,400
get in touch with us either way, even if batch 
processing is not right up your alley. We do have  

43
00:04:02,400 --> 00:04:06,880
Equidox software for ad hoc PDFs, and we also 
offer those services that I just mentioned.  

44
00:04:08,960 --> 00:04:13,040
Now, this week is a little bit different. 
So if you've joined us before for our  

45
00:04:13,040 --> 00:04:18,800
Equidox Webinar Wednesdays, you're probably used 
to me talking and demoing something about the  

46
00:04:18,800 --> 00:04:24,160
Equidox software. But this time I'm actually 
co-presenting with my friend David Freelan.  

47
00:04:24,160 --> 00:04:28,560
And he is on the Equidox team, and he is 
our lead data scientist. So I’m going to  

48
00:04:28,560 --> 00:04:32,880
turn this over to David to introduce himself 
and talk a bit more about batch processing.

49
00:04:32,880 --> 00:04:37,760
[David Freelan] Thank you, Dan. So I’m 
going to tell you a bit about myself. First,  

50
00:04:37,760 --> 00:04:43,280
I started my journey at George Mason University. I 
got my graduate degree in AI and machine learning.  

51
00:04:44,160 --> 00:04:49,040
I got recruited then by a robotic soccer 
team, and we represented the United States  

52
00:04:49,760 --> 00:04:55,440
in a competition called Robocop. I worked my 
way to team lead, and we competed in both Brazil  

53
00:04:55,440 --> 00:04:59,680
and China. I went on to publish in papers 
and deep learning and multi-agent systems.  

54
00:05:00,720 --> 00:05:07,760
After my degree, I actually had a connection at 
the Cleveland Clinic that led me to Equidox, and I  

55
00:05:07,760 --> 00:05:12,640
thought working for a company doing accessibility 
and trying to get everybody included sounded like  

56
00:05:12,640 --> 00:05:19,200
a great way to apply my skills. So I found myself 
very motivated to contribute to Equidox's goals.  

57
00:05:20,080 --> 00:05:25,040
So wrapping things around robotics has a lot 
of different components to it, but what's the  

58
00:05:25,040 --> 00:05:30,160
most relevant for us today is computer vision 
techniques, which try to figure out, let's say  

59
00:05:30,160 --> 00:05:36,640
an example for the soccer stuff where a ball or 
a robotic player is all that visual analysis of  

60
00:05:36,640 --> 00:05:41,680
what's going on on a soccer field, can actually be 
applied to analyzing a document in front of you.  

61
00:05:42,880 --> 00:05:50,400
So my initial contribution to machine learning 
and Equidox was our list and table editor.

62
00:05:52,880 --> 00:06:00,240
So when making those documents accessible, in a 
lot of situations our remediators have to specify  

63
00:06:00,240 --> 00:06:05,920
where every single row and every single column in 
a table is, as well as highlight each item in a  

64
00:06:05,920 --> 00:06:13,840
list. I can use the bullet points on this page is 
an example. A remediator would have to go through,  

65
00:06:13,840 --> 00:06:19,680
tag this whole thing as a list, tag each and every 
single item one by one. So all six items here,  

66
00:06:19,680 --> 00:06:26,240
and they'd have to make sure they tag each and 
every single delimiter. So just for this tiny  

67
00:06:26,240 --> 00:06:31,120
little list, we have a lot of different things 
we need to make sure is tagged. So we really want  

68
00:06:31,680 --> 00:06:40,480
to automate as much as this as possible so that we 
save a lot of business time. So the way we trained  

69
00:06:41,680 --> 00:06:46,640
our machine learning models, we can actually take 
that manual human process that's been done many  

70
00:06:46,640 --> 00:06:53,040
many times by our remediators, and we can teach a 
machine learning model based on those hundreds of  

71
00:06:53,040 --> 00:07:00,480
thousands of examples. We can create and we can 
then use that to consistently detect these tables  

72
00:07:00,480 --> 00:07:12,000
and lists and skip a lot of all that manual labor 
there. So while we use our internal documents  

73
00:07:12,000 --> 00:07:16,960
to train these machine learning processes, 
when we apply ourselves to batch processing  

74
00:07:17,680 --> 00:07:23,120
for a company, we're going to use your 
documents to train. We don't want a  

75
00:07:23,120 --> 00:07:27,680
general a one size fits all. We're going 
to take your documents and set ourselves  

76
00:07:29,120 --> 00:07:35,760
to identify the elements in there as accurately 
as possible. If you have a few documents out of  

77
00:07:35,760 --> 00:07:41,120
thousands or even millions of documents that look 
a bit different, we want to make sure that we use  

78
00:07:41,120 --> 00:07:47,040
every machine learning tool at our disposal to 
find those little outlier documents, which we'll  

79
00:07:47,040 --> 00:07:55,200
take a closer look at later. We want to give you a 
sneak peek on how that's done so effectively. But  

80
00:07:55,200 --> 00:08:03,520
first, let's take another look at our goal state. 
Here... (oh that's the next slide) Yeah, here's  

81
00:08:03,520 --> 00:08:12,960
our goal state here. So we want to go ahead and 
identify each of these single zones. We have an  

82
00:08:12,960 --> 00:08:21,040
icon on the top left here for your bank, we have 
a table with a header, and the table has headers  

83
00:08:21,040 --> 00:08:27,840
inside it. We got all sorts of things going on. 
Now, this is just an internal representation of  

84
00:08:27,840 --> 00:08:33,360
what's going on on our end. Don't worry, we're not 
going to be modifying your document in any way.  

85
00:08:35,600 --> 00:08:46,080
So before we... but before we train our models, 
we need to do a couple of internal steps first.  

86
00:08:46,080 --> 00:08:51,520
We're going to have to get... gather these 
documents and create all these templates,  

87
00:08:51,520 --> 00:09:00,080
and I want to show you how we do that. But first, 
let's get just a few definitions out of the way,  

88
00:09:01,680 --> 00:09:07,680
and then we can tackle a problem together. 
So artificial intelligence is just a general  

89
00:09:07,680 --> 00:09:14,640
broad scope of algorithms that are used to mimic 
human-like intelligence on any particular task.  

90
00:09:16,320 --> 00:09:22,480
More specifically, we can be interested in 
computer vision, so we're trying to mimic  

91
00:09:22,480 --> 00:09:27,840
say the visual cortex of a human and identify 
whether something is a cat or a dog or whether  

92
00:09:27,840 --> 00:09:33,760
something's a table or a list. Machine learning 
is a different subset of artificial intelligence  

93
00:09:34,480 --> 00:09:40,800
which is about teaching a model instead of like, 
hard coding. You can teach a model over time, so  

94
00:09:40,800 --> 00:09:47,360
where this intersection is, is where you're 
teaching a computer to see. So that’s kind  

95
00:09:47,360 --> 00:09:54,080
of our baseline vocabulary here. And now that 
we have that out of the way, let's go ahead  

96
00:09:54,080 --> 00:10:00,640
and try to be data scientists together for a 
second, and take a look at an example company.

97
00:10:01,200 --> 00:10:10,000
So we can take out a machine learning algorithm, 
and we can teach it how to group similar documents  

98
00:10:10,000 --> 00:10:18,640
together. So in this example on the right, each 
dot represents a single page of a document.  

99
00:10:18,640 --> 00:10:26,240
And the closer those dots are, the more similar 
that page is. So to show you a bit a better  

100
00:10:26,240 --> 00:10:33,040
example of what I mean, let's zoom in on a cluster 
here in the middle. Let's take a look at that.

101
00:10:35,200 --> 00:10:42,160
So in this cluster, we have a lot of pie charts. 
It would appear if you go from the top left,  

102
00:10:42,160 --> 00:10:47,680
the pie chart appears near the bottom... as you 
go toward the bottom right of that graph, the pie  

103
00:10:47,680 --> 00:10:54,160
chart kind of sneaks up toward the top. So just by 
looking at this cluster, zooming in, and clicking  

104
00:10:54,160 --> 00:11:00,320
on three different documents, we already have a 
really good idea. This gives our machine learning  

105
00:11:00,320 --> 00:11:07,840
algorithm and our developers to see the variation 
in these documents, just by looking at three.

106
00:11:09,920 --> 00:11:17,920
So we have the opportunity now to take a 
look at some of those points that are way  

107
00:11:17,920 --> 00:11:22,480
on the outside... that are really far away from 
all the other points. We call these sort of the  

108
00:11:22,480 --> 00:11:27,840
outlier points, and we see an interesting 
one here where the pie chart only has one  

109
00:11:27,840 --> 00:11:33,200
category. And this could be something that a 
machine learning algorithm missed, or honestly,  

110
00:11:34,080 --> 00:11:41,280
a human who is handwriting a template could 
totally miss out on this situation. If there's  

111
00:11:41,280 --> 00:11:45,920
thousands or millions of documents, this is 
just potentially one in a million that could  

112
00:11:45,920 --> 00:11:50,240
be hard to find. So our machine learning 
algorithm was able to find that outlier,  

113
00:11:50,240 --> 00:11:55,440
and handle that. All right so there's one 
other thing I actually want to note on this  

114
00:11:55,440 --> 00:12:01,760
page. There's a 100 there that's right on top of 
the pie chart. That can actually kind of blend in  

115
00:12:01,760 --> 00:12:06,880
with that paragraph above it. So again, without 
noticing these sorts of things, and going through  

116
00:12:06,880 --> 00:12:12,000
this initial machine learning process, there's 
a good chance we would miss that 100 percent.  

117
00:12:14,240 --> 00:12:19,040
And missing something means that someone 
isn't included. So we want to make sure we get  

118
00:12:19,040 --> 00:12:27,680
everything. Let's take a look at this real fast. 
So this cluster we’ll be zoomed into has a lot of  

119
00:12:27,680 --> 00:12:34,480
letters on it. I wanted to show here that while we 
obviously doing a lot more than a fixed overlay,  

120
00:12:34,480 --> 00:12:39,840
in this cluster, there are some static 
elements that say, the logo on the top,  

121
00:12:40,400 --> 00:12:45,600
and the address, that are all in generally the 
same location. We can actually take advantage  

122
00:12:45,600 --> 00:12:52,480
of that and make sure that our algorithm focuses 
on getting these sorts of variations here. As we  

123
00:12:52,480 --> 00:12:58,080
look at these three individual points, we can 
see that these paragraphs can vary in length.  

124
00:12:58,080 --> 00:13:04,480
Some of this header can vary in size, so we 
need to make sure that we account for that.  

125
00:13:05,520 --> 00:13:10,480
We sort of have a mix of a static and 
dynamic thing going on in this cluster.

126
00:13:13,600 --> 00:13:18,240
Okay so in these three clusters, these clusters  

127
00:13:18,240 --> 00:13:22,880
appear to be closer together because 
they have a sort of similar style, but  

128
00:13:26,000 --> 00:13:31,840
each cluster has a different number of 
columns. So what I wanted to show you all here,  

129
00:13:31,840 --> 00:13:37,280
is how each document even though they have a 
differing number of columns, they have a lot of  

130
00:13:37,280 --> 00:13:43,760
similar features. So actually instead of building 
three different templates, one for each cluster,  

131
00:13:43,760 --> 00:13:52,240
I could actually split up the document by their 
columns and create one template that can handle  

132
00:13:52,240 --> 00:13:59,440
columns. And I could sort of combine all these 
clusters together. And by graphing all these out,  

133
00:13:59,440 --> 00:14:05,840
gave me a good view of how these differ and 
how these documents are so similar to each  

134
00:14:05,840 --> 00:14:11,280
other. So machine learning algorithm really really 
helped us learn a bit more about these documents.

135
00:14:14,800 --> 00:14:22,960
So the last kind of document I wanted to show you 
today is where the tables vary in their structure.  

136
00:14:22,960 --> 00:14:30,800
So at the top left here, we have very few cells 
compared to, if we go all the way down on the  

137
00:14:30,800 --> 00:14:39,200
graph, on the lower right we could see a lot 
of very dense cells. So we want... our machine  

138
00:14:39,200 --> 00:14:44,560
learning algorithm here is informing us that this 
is primarily how the document is going to change.  

139
00:14:45,200 --> 00:14:52,080
That the amount of cells in this table. So we want 
to make sure that our machine learning algorithm  

140
00:14:52,080 --> 00:14:57,680
not only identifies these PDF elements, but 
is able to tag them correctly despite these  

141
00:14:57,680 --> 00:15:04,240
variations. So we have our own table algorithms 
to make sure that we handle this appropriately.

142
00:15:06,720 --> 00:15:11,280
And let's again let's take a look at another 
outlier point because this could be really  

143
00:15:11,280 --> 00:15:18,720
important example. So in the top left, this 
could be a bug on this company's end, or it  

144
00:15:18,720 --> 00:15:23,760
could be something that's an outlier that we need 
to make sure that we can handle. But there's a one  

145
00:15:23,760 --> 00:15:29,120
on the top left that's completely missing a table. 
There's no cells at all. So this is another thing  

146
00:15:29,120 --> 00:15:34,560
that if those... if this cluster had a million 
documents in it without something like this  

147
00:15:34,560 --> 00:15:39,840
algorithm you might never find it looking for it 
by hand. So this is a case where we could either  

148
00:15:39,840 --> 00:15:47,440
alert you that this customer might have an issue, 
or we can just keep that in mind to ourselves. If  

149
00:15:47,440 --> 00:15:56,320
you say it's not a problem, and make sure that we 
are handling that edge case appropriately. So once  

150
00:15:56,320 --> 00:16:02,000
we've done all of that we can build ourselves a 
lovely template. This is the same thing we saw  

151
00:16:02,000 --> 00:16:07,840
on the last slide. Sort of a reminder of what 
our goal state i. We've now been able to show you  

152
00:16:09,280 --> 00:16:15,200
how we can identify all these different elements 
on the page. And this is what internally (again  

153
00:16:15,200 --> 00:16:21,840
we're not marking your documents) how internally 
all these things will be tagged. They will make  

154
00:16:21,840 --> 00:16:27,680
sure we apply them correctly so that their tags 
are compliant accessible and usable. Everything  

155
00:16:27,680 --> 00:16:36,320
you'd want to see. So now you might be wondering, 
cool we’ve made all these templates, we've done  

156
00:16:36,320 --> 00:16:42,240
all this fancy machine learning stuff, how are 
we going to get these documents to your company?  

157
00:16:42,240 --> 00:16:46,320
You have documents, you want them remediated by 
our batch processing system… How's that going  

158
00:16:46,880 --> 00:16:55,760
to be, going to work for you? So what we want to 
supply for your developers is an interface where  

159
00:16:55,760 --> 00:17:02,640
you can call a rest API. And with that rest 
API... that rest API on our end is going to  

160
00:17:02,640 --> 00:17:08,960
interact with either a local server or a cloud 
server that can be scaled up and down. Whether  

161
00:17:08,960 --> 00:17:18,640
it's local- or cloud-based is your preference. 
So once we have that PDF, we can go along  

162
00:17:18,640 --> 00:17:25,760
and forward that to the machine learning algorithm 
that we have just described in great detail. And  

163
00:17:26,880 --> 00:17:30,400
after the machine learning has identified 
the templates, and tagged them all correctly,  

164
00:17:32,560 --> 00:17:40,080
we can send that simply right back to you through 
the same rest API. And you have yourselves a  

165
00:17:40,080 --> 00:17:45,840
tagged and accessible document for your end-users.  

166
00:17:50,640 --> 00:17:55,440
Okay, so that concludes this. Dan, if 
you can take it away, we have a few FAQs. 

167
00:17:55,440 --> 00:18:02,880
[Dan Tuleta] Sure thank you David. So when we are 
talking about batch processing with our prospects  

168
00:18:02,880 --> 00:18:08,560
and existing clients, there's a lot of common 
questions that come up. So what we've tried to  

169
00:18:08,560 --> 00:18:14,240
do within this presentation is to just compile 
a list of these FAQs. And we're going to walk  

170
00:18:14,240 --> 00:18:19,200
through a few of those common questions that we 
feel that many of you on this call are probably  

171
00:18:19,200 --> 00:18:22,880
wondering right now. So David if you want to 
jump to the next slide and we can get started.  

172
00:18:24,720 --> 00:18:32,160
Okay so David, is this really accessible? [David] 
Yes, yes it is. So it's not just my expertise  

173
00:18:32,160 --> 00:18:39,120
that's going in into this. I'm not a one-man 
show. We have our Director of Accessibility and  

174
00:18:39,120 --> 00:18:44,080
his entire team that's going to be going through 
with our developers. And all the machine learning  

175
00:18:44,080 --> 00:18:48,480
process here to make sure that 
we're complying with everything.  

176
00:18:50,480 --> 00:18:58,320
We don't want to leave anybody out. [Dan] Right. 
Our PDF remediation team is full of PDF experts.  

177
00:18:58,320 --> 00:19:02,480
When we send documents back to clients 
as part of our PDF remediation services,  

178
00:19:03,040 --> 00:19:08,320
everything goes through a multi-step validation 
process. So every document, every template that  

179
00:19:08,320 --> 00:19:13,600
we're designing for you, is going to go through 
rigorous validation and this is not just trying  

180
00:19:13,600 --> 00:19:18,960
to trick the accessibility checkers into saying 
that this document is accessible. We are actually  

181
00:19:18,960 --> 00:19:23,840
building a fully tagged, fully accessible, 
fully usable document for your clients.  

182
00:19:24,640 --> 00:19:29,600
Now again this is not an overlay so these are not 
static documents that have to be exactly the same  

183
00:19:29,600 --> 00:19:34,800
way, and then you just place or copy and paste 
the exact same template onto every document.  

184
00:19:35,680 --> 00:19:38,960
These documents might have minor 
variances, as David has explained.  

185
00:19:39,520 --> 00:19:43,760
Even if that's the case our machine 
learning algorithms are able to detect that  

186
00:19:43,760 --> 00:19:48,000
and every document will be tagged in a unique 
way to make sure that it is fully usable.

187
00:19:50,880 --> 00:19:54,160
So David, a lot of people are wondering 
about these templates. You know,  

188
00:19:54,160 --> 00:19:58,400
how many templates can you have? What if 
they need to change? And if they do need to  

189
00:19:58,400 --> 00:20:06,720
change how long do these changes take to apply?
[David] Yeah so the number of templates is very  

190
00:20:06,720 --> 00:20:12,480
fluid. I wouldn't necessarily be too concerned 
about our number, because templates can vary  

191
00:20:12,480 --> 00:20:17,680
in difficulty. So that's the thing we can do 
on a case-by-case basis. If they change it's,  

192
00:20:17,680 --> 00:20:24,480
again it depends on how complex the changes. I 
imagine a lot of changes can vary in difficulty.  

193
00:20:24,480 --> 00:20:30,800
So that's again, unfortunately, one of those 
things we do tackle on a case-by-case basis. And  

194
00:20:31,920 --> 00:20:37,360
the length of time is again, it's one of those 
unsatisfying answers potentially. But yes we  

195
00:20:37,360 --> 00:20:41,280
want to make sure we're working with you, and 
the length of time it takes us to complete  

196
00:20:41,280 --> 00:20:49,920
the projects is going to depend on your needs.
[Dan] Great. All right. That's just one thing  

197
00:20:49,920 --> 00:20:54,160
to keep in mind, is that all of these are 
custom solutions. So that's going to be sort  

198
00:20:54,160 --> 00:20:59,120
of a theme throughout the FAQs. We really 
do need to engage with you and your team  

199
00:20:59,680 --> 00:21:04,800
on a one-to-one basis and understand your 
documents, your templates, and all of  

200
00:21:04,800 --> 00:21:09,280
your needs to come up with a custom solution 
that will fit for for all of your documents.

201
00:21:12,560 --> 00:21:14,880
So David, can the system  

202
00:21:14,880 --> 00:21:19,840
run multiple templates simultaneously?
[David] Yes that's really not a problem.  

203
00:21:19,840 --> 00:21:25,280
Running an arbitrary amount of templates is not 
an issue for us. We just need enough processing  

204
00:21:25,280 --> 00:21:30,240
power to do it. And we don't think it's 
going to take that much processing power to  

205
00:21:30,240 --> 00:21:38,800
do. So it's really, really not a problem for us.
[Dan] Great, all right. And so David, of course,  

206
00:21:38,800 --> 00:21:43,920
a very important question. Is this process fully 
automated? Or does it need any sort of human  

207
00:21:43,920 --> 00:21:49,200
hand-holding or babysitting on a day-to-day basis?
[David] It is absolutely automated. It's  

208
00:21:49,200 --> 00:21:54,080
just a matter of setting that those initial 
conversations and creating those templates.  

209
00:21:54,080 --> 00:22:00,000
And from there it should just go smoothly.
[Dan] Great. Once we've had those  

210
00:22:00,000 --> 00:22:02,640
initial discovery calls. and we 
better understand your templates,  

211
00:22:03,280 --> 00:22:05,760
your documents, and we can 
develop those templates,  

212
00:22:05,760 --> 00:22:11,840
we can take the human involvement completely out 
of the equation and it all just works like magic.

213
00:22:15,040 --> 00:22:22,880
So another question that we're often asked is, 
should this be an on-prem or a cloud-based type  

214
00:22:22,880 --> 00:22:30,880
of solution? So in talking with our developers, 
both options are available. So we can set up this  

215
00:22:30,880 --> 00:22:37,840
system either in the cloud or locally installed 
on your servers. Our developers have recommended  

216
00:22:37,840 --> 00:22:44,160
for this type of system that on-prem is going to 
be faster because you won't have data that has  

217
00:22:44,160 --> 00:22:51,040
to ping back and forth from the cloud going both 
directions. So everything can stay local into in  

218
00:22:51,040 --> 00:22:57,600
your own environment. And also this might help you 
meet some of your internal security requirements.  

219
00:22:57,600 --> 00:23:03,200
So a lot of these documents might contain client 
data, or information about specific people,  

220
00:23:03,200 --> 00:23:09,520
or account numbers, for example. So there might 
be a lot of very strict and rigorous internal  

221
00:23:09,520 --> 00:23:15,040
IT security measures that you have to adhere to. 
So keeping it locally installed on your server in  

222
00:23:15,040 --> 00:23:19,200
your environment might be a better option. But 
if you are interested in having a conversation  

223
00:23:19,200 --> 00:23:24,400
about how to install this on the cloud, that's 
absolutely fine. We can support either option.  

224
00:23:27,520 --> 00:23:31,920
Now David, what kind of volume and 
scalability are supported here? 

225
00:23:31,920 --> 00:23:38,720
[David] Yeah, it's it's arbitrarily large. You 
can have a billion documents going through and  

226
00:23:38,720 --> 00:23:44,320
we can handle it. You just need the hardware 
to do it. But we don't exactly expect you to  

227
00:23:44,320 --> 00:23:49,520
have to rent a supercomputer or anything 
to do that sort of thing. But it's again,  

228
00:23:49,520 --> 00:23:54,960
one of those case-by-case things. So 
we'll work with you to determine on  

229
00:23:54,960 --> 00:24:00,080
exactly how big of a machine you might need.
[Dan] Right there can be a lot of variance  

230
00:24:00,080 --> 00:24:05,600
between the size of the client. Whether you're a 
small credit union, or a small utility company,  

231
00:24:05,600 --> 00:24:10,800
or a Fortune 500 company. You could be talking 
about thousands of documents a month, or maybe  

232
00:24:10,800 --> 00:24:15,040
millions of documents per month. But we can scale 
accordingly. It really is just dependent on the  

233
00:24:15,040 --> 00:24:19,200
hardware that you are trying to run this through. 
But that again is a conversation that we can have  

234
00:24:19,760 --> 00:24:22,800
offline once we're having a one-to-one 
discussion with you and your team.

235
00:24:25,840 --> 00:24:35,040
Now another question that we're often asked is, 
is it secure? And yes it is secure. Whether you  

236
00:24:35,040 --> 00:24:43,200
go with the cloud option or the on-prem, it is in 
fact a very secure system. So many people might  

237
00:24:43,200 --> 00:24:48,080
opt for the on-prem install just for that extra 
layer of security, keeping everything installed  

238
00:24:48,080 --> 00:24:54,160
locally on your own environment to adhere to all 
of your internal IT security rules and regulation.  

239
00:24:54,800 --> 00:25:01,280
But just keep in mind that Equidox is a member of 
the GSA schedule, California’s CMAS, the Canadian  

240
00:25:01,280 --> 00:25:06,464
SLSA, and we've also passed our ISO certification 
as well. So keep those in mind when you  

241
00:25:08,160 --> 00:25:13,920
take everything away from this presentation. 
That this is in fact a secure solution, and we  

242
00:25:13,920 --> 00:25:18,480
can have a more detailed discussion about any of 
your specific requirements when the time comes.

243
00:25:25,120 --> 00:25:30,160
I didn't even notice you changed the slide, David. 
Thank you so what does it cost? This is of course  

244
00:25:30,160 --> 00:25:34,400
the million dollar question. So everyone wants 
to know what will this cost. And it really can  

245
00:25:34,400 --> 00:25:40,400
vary depending on the complexity of the templates, 
the number of templates that you have, the volume,  

246
00:25:40,400 --> 00:25:44,800
and the hardware that you are running it on. So 
there's a number of factors that go into the price  

247
00:25:45,520 --> 00:25:50,000
and we know that this has to scale up and 
down based on a lot of different factors. So  

248
00:25:50,000 --> 00:25:54,640
what we want to do is of course have a 
conversation with you to talk about your  

249
00:25:54,640 --> 00:25:59,760
exact needs… talk about your documents... 
better understand the situation for your  

250
00:25:59,760 --> 00:26:04,720
organization. And then we will come up with a 
custom pricing solution that will work for both  

251
00:26:04,720 --> 00:26:10,800
parties. So this has to be kind of discussed 
on a one-to-one basis. There is no just one  

252
00:26:10,800 --> 00:26:15,280
size or one price fits all type of solution 
here. Everything is custom keep that in mind.

253
00:26:18,400 --> 00:26:25,200
So in summary that kind of wraps up our 
FAQ. So in summary, Equidox batch processing  

254
00:26:25,840 --> 00:26:30,000
is a customized service for our clients 
to produce fully accessible documents in  

255
00:26:30,000 --> 00:26:36,240
a fully automated way. Without any day-to-day 
human involvement. So this allows for on-demand  

256
00:26:36,240 --> 00:26:40,560
accessible PDF creation. So that when 
any of your customers download a PDF,  

257
00:26:40,560 --> 00:26:46,080
whether it be their bank statement, or a utility 
bill, or any sort of templatized document,  

258
00:26:46,080 --> 00:26:50,720
Equidox batch processing will automatically 
produce that document in an accessible format.  

259
00:26:51,440 --> 00:26:56,880
So our batch processing solution is a 
fully secure, fully scalable solution  

260
00:26:56,880 --> 00:27:01,200
to meet your needs. Whether you're producing 
a few thousand or a few million documents,  

261
00:27:01,200 --> 00:27:07,520
this system will take the burden of page-by-page 
remediation out of your hands. It will also  

262
00:27:07,520 --> 00:27:11,520
ultimately ensure that there is inclusion 
for all of your clients, and it will mitigate  

263
00:27:11,520 --> 00:27:15,840
your legal risk of complying with all of the 
accessibility requirements for your organization.

264
00:27:19,200 --> 00:27:22,640
So that is going to conclude 
our presentation today. As  

265
00:27:22,640 --> 00:27:33,840
always please feel free to reach out to us 
either through our website www.Equidox.com  

266
00:27:40,720 --> 00:27:44,000
about Equidox in general. You can definitely surf  

267
00:27:44,000 --> 00:27:47,840
through our website and of course, 
we are very active on… [silence]

268
00:27:51,880 --> 00:27:52,880
[Tammy]  

269
00:27:52,880 --> 00:27:56,640
So Dan appears to be having some connection 
problems. So we just want to share that we are  

270
00:27:57,360 --> 00:28:02,320
very active on social media. You can find 
us on LinkedIn or Facebook or Twitter.  

271
00:28:03,120 --> 00:28:08,080
You can follow our business pages there and get 
any updates that you might need or be interested  

272
00:28:08,080 --> 00:28:17,040
in. And we're really always trying to keep new 
material going. And you'll see announcements  

273
00:28:17,040 --> 00:28:21,280
for our next webinar there. I really want 
to thank you for showing up today David. 

274
00:28:21,280 --> 00:28:22,720
[David] My pleasure.
[Tammy] That was a really  

275
00:28:23,280 --> 00:28:29,160
cool explanation. Is there anything you'd like 
to say to the crowd before we close up today? 

276
00:28:29,160 --> 00:28:35,120
[Dan] I think I dropped my internet. I do 
apologize if I was in the middle of a sentence.  

277
00:28:35,120 --> 00:28:39,200
But thank you for picking things up Tammy.
[Tammy] Yeah no problem. We were just giving  

278
00:28:39,200 --> 00:28:45,840
David a moment to say his last hurrah here.
[David] Yeah I’d like to thank you all for  

279
00:28:45,840 --> 00:28:49,680
listening. And I hope you have a 
good reassurance that we are doing  

280
00:28:49,680 --> 00:28:55,200
our best to make sure everything is remediated as 
accurately as possible. And we are not slacking!

281
00:28:56,280 --> 00:29:11,840
[Dan] Great thank you, everyone!
[David] Thank you!