WEBVTT

00:00:00.000 --> 00:00:06.240
You're welcome. Have a great day, 
everyone. All right, bye-bye. Thank you,  

00:00:06.240 --> 00:00:12.900
guys. All right, guys, we're going to 
move in now to our next presentation.

00:00:15.360 --> 00:00:22.920
We're going to have our next presentation by 
Miss Dana Chisnell, Acting Executive Director  

00:00:22.920 --> 00:00:29.820
for Customer Experience at the Department of 
Homeland Security. Hello, Dana, can you hear me?

00:00:34.860 --> 00:00:36.120
>> Yes, I can hear you.

00:00:36.120 --> 00:00:39.420
>> All right, I can see you. I just want to 
make sure that you can hear me. Sometimes  

00:00:39.420 --> 00:00:44.220
that transition, you know, between presenters, 
you just never know what happens. Just want to  

00:00:44.220 --> 00:00:49.980
make sure everything is working smoothly. Very 
good. Seeing you again today, I appreciate that.

00:00:49.980 --> 00:00:56.760
>> Likewise, thank you for introducing 
me, as you do that ritual where I ask,  

00:00:56.760 --> 00:01:00.660
are you seeing my slides for the 
sighted people in the audience?

00:01:03.260 --> 00:01:07.920
>> And I can't hear anyone, so 
that's right. So if you guys can  

00:01:07.920 --> 00:01:11.580
see the slides and if you can hear Miss 
Chisnell, please put a one in the chat.

00:01:16.680 --> 00:01:20.400
Okay, the ones are coming in hot and 
heavy again. Looks like we are good to go.

00:01:20.400 --> 00:01:26.820
>> Excellent! Thank you so much. Hello 
everyone, I'm Dana Chisnell. I'm Acting  

00:01:26.820 --> 00:01:32.040
Executive Director for Customer Experience 
at the Department of Homeland Security. My  

00:01:32.040 --> 00:01:39.060
pronouns are she/her. I'm an older white 
woman with short graying brown hair and  

00:01:39.060 --> 00:01:44.400
I wear brown square glasses. Today, I'm 
wearing a red speckled blouse with a bow.  

00:01:45.000 --> 00:01:49.440
I am a person who has not been willing 
to admit that I may be disabled.

00:01:50.100 --> 00:01:54.720
The last panel was really fantastic, and I 
can't thank the speakers enough for sharing  

00:01:54.720 --> 00:02:02.100
their knowledge and experience and being out 
with their disabilities. I'm really honored  

00:02:02.100 --> 00:02:08.460
to get to speak with you all today and welcome our 
employees and contractors and our agency partners.

00:02:10.260 --> 00:02:16.440
What I want to talk about is how an inclusive 
environment that centers experiences means  

00:02:16.440 --> 00:02:22.380
creating an accessible environment, and I loved 
Annette's metaphor, the best metaphor of the day,  

00:02:22.380 --> 00:02:27.240
the blueberry muffin thing. I'm 
stealing that, it's awesome.

00:02:29.400 --> 00:02:35.220
So, the Department of Homeland Security, 
with its wide range of operating components,  

00:02:35.220 --> 00:02:40.980
has about 1 billion interactions, in 
person, with the public every year.  

00:02:41.880 --> 00:02:48.060
The numbers and the scale are staggering, 
really. This slide shows a few examples,  

00:02:48.060 --> 00:02:55.020
for, um, one, roughly 25 million people 
survive natural disasters every year.  

00:02:56.580 --> 00:03:07.380
The Customs and Border Protection, um, offices 
and ports of Entry process around 492,000  

00:03:07.380 --> 00:03:19.500
passengers every day. USCIS conducted about 
9,900 Refugee interviews in 2021. Employers  

00:03:19.500 --> 00:03:31.200
verified around 225,759 people through USCIS' 
e-verify Service, uh, just in FY22.

00:03:34.380 --> 00:03:41.160
750 million people travel through 
U.S airports every year, and TSA  

00:03:41.160 --> 00:03:47.040
screens about 8 million Travelers 
per week under TSA pre-check alone.

00:03:49.560 --> 00:03:55.140
So, um, customers at DHS span from 
disaster survivors to refugees to  

00:03:55.140 --> 00:04:01.560
domestic and international Travelers, along with 
employers and non-citizens, and so many others.  

00:04:01.560 --> 00:04:11.640
For the customer experience team in DHS OCIO, 
we also see DHS employees as our customers  

00:04:11.640 --> 00:04:15.360
because we know that the experience 
that you have as an employee  

00:04:16.380 --> 00:04:22.980
leaks out into the customer experience for the 
public. That if you are comfortable and have  

00:04:22.980 --> 00:04:29.100
what you need to do your work, um, you're going 
to deliver a better experience to the public.

00:04:30.780 --> 00:04:35.940
A key element of customer experience design is 
understanding who your customers are and what  

00:04:35.940 --> 00:04:41.880
their needs are. Their needs also come 
with context, right? So, like, what are  

00:04:41.880 --> 00:04:49.860
they trying to achieve and, and why? What's the 
context they're in and, um, when they're trying  

00:04:49.860 --> 00:04:55.020
to achieve their goals? Is it stressful, is it 
noisy, do they have everything that they need,  

00:04:55.980 --> 00:05:02.280
what do they know already about the service or the 
process, and what they will have to learn while  

00:05:02.280 --> 00:05:11.220
they're trying to reach their own goals? How much 
do they know, um, sorry, how are their skills in  

00:05:11.220 --> 00:05:17.040
understanding how government is organized? This is 
the thing that we force people to do, right? Like,  

00:05:17.040 --> 00:05:24.480
you have to know which agency to go to for which 
things, which, um, is not always intuitive.

00:05:26.340 --> 00:05:30.600
What physical, cognitive, or intellectual 
disabilities might someone have?  

00:05:31.800 --> 00:05:38.400
What's the literacy level that they have, and 
what amount of English do they actually have  

00:05:39.060 --> 00:05:43.200
to understand all of that? We borrow 
heavily from a set of methods from humans,  

00:05:46.800 --> 00:05:52.260
design, instead of real people, we make 
interactions accessible and equitable.

00:05:55.620 --> 00:06:01.200
So, looking at the shared goals of accessibility 
and customer experience across DHS, um,  

00:06:01.200 --> 00:06:07.380
is what I'm here to talk to you about today. So, 
integrating accessibility best practices during  

00:06:07.380 --> 00:06:12.900
design, um, putting, mixing the blueberries 
in the muffin, not leaving them on top.  

00:06:14.040 --> 00:06:17.580
Um, integrating accessibility 
practices during design and  

00:06:17.580 --> 00:06:21.780
development will enhance the customer 
experience for everyone. Accessibility  

00:06:21.780 --> 00:06:25.920
further improves customer engagement 
by removing barriers to content.

00:06:26.700 --> 00:06:33.240
I'm going to cover three areas of shared goals 
between CX and accessibility and highlight an  

00:06:33.240 --> 00:06:39.120
example of each of the goals currently in 
our work streams. Lots of these overlap  

00:06:39.120 --> 00:06:43.860
and the examples could be interchangeable 
with any of these areas to some degree.

00:06:45.780 --> 00:06:52.740
So, first, we want to talk about inclusive design. 
Um, as I mentioned, a lot of people are very  

00:06:52.740 --> 00:06:57.360
stressed when they interact with the Department of 
Homeland Security, to think they're high, right?  

00:06:57.960 --> 00:07:06.540
Um, people need benefits often in a hurry and, 
or they need decisions quickly for a variety  

00:07:06.540 --> 00:07:13.440
of reasons. So, there's an added cognitive load 
that disables people even further than they might  

00:07:13.440 --> 00:07:21.540
already be. Many people DHS interacts with have 
low English as well. The department is getting  

00:07:21.540 --> 00:07:28.260
better at language access, but there's still a 
considerable way to go on this. But we don't talk  

00:07:28.260 --> 00:07:36.540
about enough and don't design services well for is 
low literacy. About 54% of American adults read at  

00:07:36.540 --> 00:07:44.400
or below the sixth-grade level. This covers a lot 
of territory, uh, a lot of humans. But when you  

00:07:44.400 --> 00:07:51.780
add in refugees and asylees who are not citizens 
yet, and people coming through parole programs,  

00:07:51.780 --> 00:08:00.900
uh, they're very likely to have low literacy as 
well. Um, but DHS tends to default to delivering  

00:08:00.900 --> 00:08:07.620
information in print that is challenging for 
lots of people to comprehend, even for people  

00:08:07.620 --> 00:08:13.800
with high literacy. So, we're asking the question, 
why does this happen and what can we do about it?

00:08:13.800 --> 00:08:21.780
So, as CX and OAST partner closely, uh, 
the result will be increased capacity to  

00:08:21.780 --> 00:08:28.020
better address intersections of identity. So, 
think about structurally minoritized and, uh,  

00:08:28.020 --> 00:08:32.340
or racialized individuals who 
also live with disabilities,  

00:08:32.340 --> 00:08:39.720
people with low English proficiency, uh, and 
more. So, uh, we're going to be realigning  

00:08:40.320 --> 00:08:47.220
CX and Accessibility together to increase all 
that capacity, and I'm very excited about that.

00:08:49.380 --> 00:08:56.280
So, um, here's an example. Uh, we're talking 
about accessible plain language. The CX team  

00:08:56.280 --> 00:09:03.120
in OCIO is spinning up a new accessible plain 
language practice, um, and we're doing that  

00:09:03.120 --> 00:09:10.320
by establishing a dedicated group within a 
vision to provide accessible, anti-racist,  

00:09:10.320 --> 00:09:15.960
culturally responsive, equitable, and 
inclusive plain language services.  

00:09:16.980 --> 00:09:26.040
For them, we are also planning to do this 
by delivering on the secretary's priority  

00:09:26.040 --> 00:09:30.060
on improving equity and indeed just 
programs, policies, and initiatives,  

00:09:30.060 --> 00:09:36.180
including through increasing language access 
by reviewing complaint processes, as well as  

00:09:36.180 --> 00:09:43.920
the I-Speak materials to ensure resources 
are accessible, meaningful, comprehensible,  

00:09:44.700 --> 00:09:52.200
um, and meets the needs of low English proficiency 
and low literacy folks. We're going to champion  

00:09:52.200 --> 00:09:59.880
plain language and language access as core skills 
that everyone across DHS should take into account.

00:10:04.920 --> 00:10:12.240
As you can imagine, taking initiatives like 
this, um, mean that there's a lot of culture  

00:10:12.240 --> 00:10:17.160
change involved. So, part of the culture 
change is addressing this idea of customer,  

00:10:17.160 --> 00:10:24.900
um, the CX team reframes this to talking about 
human-centered or user-centered, um, Services,  

00:10:24.900 --> 00:10:31.980
which makes the goal more relatable. So, you may 
have noticed that at the beginning of the, um,  

00:10:31.980 --> 00:10:38.580
presentation, I talked about customers and, um, 
named the different kinds. That's where we focus,  

00:10:39.900 --> 00:10:46.260
um, so we asked the question, um, if we do a 
great job with this service or this product,  

00:10:46.260 --> 00:10:51.540
whose lives are improved and in 
what ways. That conversation also  

00:10:51.540 --> 00:10:56.760
supports talking about outcomes rather than 
outputs, so how we approach this culture  

00:10:56.760 --> 00:11:02.040
change is by leveraging partnerships, working 
together with internal partners within DHS,  

00:11:02.040 --> 00:11:08.520
and collaborating with community organizations 
to facilitate these, um, human-centered outcomes.

00:11:11.700 --> 00:11:18.240
So, um, partnering with NGOs helps us 
to understand the needs of non-citizens,  

00:11:18.240 --> 00:11:23.880
for example, um, there are a couple of notable 
examples projects that I can talk about. One  

00:11:23.880 --> 00:11:29.160
of them is that we collaborated with Customs and 
Border Protection to improve the user experience  

00:11:29.160 --> 00:11:36.840
for the CBP One mobile application that people 
are using to submit Advanced information, um,  

00:11:36.840 --> 00:11:46.140
to a Port of Entry to get an appointment to, um, 
be screened for parole. We get to partner with  

00:11:46.140 --> 00:11:51.480
other parts of the organization as well, like ICE, 
which is working on Plain language observation  

00:11:51.480 --> 00:11:57.240
for non-citizens who are apprehended at the 
border, so they know what process they're in  

00:11:57.240 --> 00:12:06.060
and what their obligations are under whatever, um, 
immigration status that they have. So, we partner  

00:12:06.060 --> 00:12:12.720
a lot with internal offices in DHS, such as the 
Civil Rights and Civil Liberties group, I.T.,  

00:12:12.720 --> 00:12:18.840
and many more to create a culture that centers 
accessibility, um, and customer experience.  

00:12:19.440 --> 00:12:25.680
Through the DHS paperwork reduction act burden 
reduction initiative, that was a lot of words, and  

00:12:25.680 --> 00:12:32.700
I'm sorry for anybody who's doing sign language 
interpretation or, um, if the transcription didn't  

00:12:32.700 --> 00:12:40.560
come through very well. Through the DHS paperwork 
reduction act burden reduction initiative,  

00:12:42.240 --> 00:12:49.140
CX partnered with components to implement 
various initiatives using human-centered burden  

00:12:49.140 --> 00:12:56.460
reduction methods to build more equitable customer 
experiences for both employees and for the public.

00:13:01.380 --> 00:13:09.240
Third, um, a lot of, um, the modernization that is 
happening that many of you are probably involved  

00:13:09.240 --> 00:13:16.380
with across the Department of Homeland Security 
isn't public-facing. Um, it's improving systems  

00:13:16.380 --> 00:13:22.620
for government, um, personnel. Um, it's important 
because the experience they have leaks into the  

00:13:22.620 --> 00:13:28.620
experience that our public customers have. 
Additionally, um, accessibility must be a  

00:13:28.620 --> 00:13:36.240
core consideration from the start of the project. 
Right now, um, for example, DHS treats video with  

00:13:36.240 --> 00:13:44.760
captions and audio and language in other languages 
besides English as exceptions, mostly because, um,  

00:13:44.760 --> 00:13:52.080
I believe the organization doesn't have good tools 
or resources. There aren't enough people who have  

00:13:52.080 --> 00:14:00.240
the skills to implement these things. Um, we're 
going from policy directives to or guidance to  

00:14:00.240 --> 00:14:08.160
plain language printed public information and to 
other delivery channels, so we need to do better  

00:14:08.160 --> 00:14:13.980
on these things. And so I call on everyone who 
is in this virtual room today to think about  

00:14:13.980 --> 00:14:23.640
how the information that you're creating that 
ultimately goes out to the public, um, how you  

00:14:23.640 --> 00:14:31.860
can make sure that that's transportable to lots 
of other delivery methods and reception methods  

00:14:33.780 --> 00:14:40.500
by incorporating best practices like usability 
testing. We can save customers a significant  

00:14:40.500 --> 00:14:47.820
amount of time and energy because we'll have a 
much better, um, idea of what their needs are.

00:14:52.260 --> 00:14:58.860
So, um, I mentioned that we were working with ICE 
on, um, informed compliance for non-citizens who  

00:14:58.860 --> 00:15:05.280
are in the country. Um, the Secretary's priority, 
Secretary Mayorkas, has created a priority that  

00:15:05.280 --> 00:15:12.240
includes, um, establishing and implementing a 
research-driven framework for communications with  

00:15:12.240 --> 00:15:19.080
non-citizens who are apprehended at the Southwest 
border that increases compliance with proceedings  

00:15:19.080 --> 00:15:30.900
and language access. And this is meant to stretch 
across DHS, so from our website at dhs.gov to CBP  

00:15:30.900 --> 00:15:38.580
One, the app from Customs and Border Protection, 
the ICE non-citizen website and portal for  

00:15:41.280 --> 00:15:52.080
scheduling updates, and related websites and other 
information at USCIS and the Department of Justice  

00:15:52.080 --> 00:15:58.800
because this process stretches into that. We 
want to cover all of those things digitally as  

00:15:58.800 --> 00:16:04.140
well as through other kinds of messaging and 
any kinds of fact sheets that might go out.  

00:16:05.820 --> 00:16:16.140
So, um, the DHS CX team has taken on several 
multilateral initiatives aimed at simplifying  

00:16:16.140 --> 00:16:22.860
the experience that non-citizens are having as 
they navigate the immigration system. Right now,  

00:16:22.860 --> 00:16:27.060
it looks like you're interacting with four 
different governments, not different parts  

00:16:27.060 --> 00:16:34.980
of the same government. And so, we would like 
to take the first steps toward rectifying that,  

00:16:34.980 --> 00:16:41.760
making it easier to, um, to make sure 
that you're meeting your responsibilities.

00:16:43.860 --> 00:16:51.000
The goal is improving case management, 
um, and compliance, really. So, our team  

00:16:51.000 --> 00:16:59.400
is partnering with ICE to deliver designs for 
a website, um, as a first step. And then with  

00:16:59.400 --> 00:17:04.500
Customs and Border Protection, ICE, uh, 
U.S. Citizenship and Immigration Services,  

00:17:04.500 --> 00:17:12.120
and various working groups that are involved in 
the southwestern border, we've pulled together  

00:17:12.120 --> 00:17:18.540
a little working group to agree on an approach, 
um, to ensure that migrants and non-citizens, uh,  

00:17:18.540 --> 00:17:25.020
can get the information they need, understand 
what they find, and take actions to comply.

00:17:29.520 --> 00:17:34.440
So, by creating equitable, accessible, 
and human-centered services,  

00:17:35.460 --> 00:17:39.120
we build a strong foundation 
of trust in the agency.  

00:17:40.260 --> 00:17:47.640
So, we'll know that we're successful when we're 
all asking, whenever we have a decision to make,  

00:17:48.180 --> 00:17:54.060
what are the outcomes that we want for the people 
we're serving, and what's the experience we want  

00:17:54.060 --> 00:18:01.680
them to have as they're reaching that? 
That is where I want to finish out today,  

00:18:01.680 --> 00:18:07.340
and now I'll be happy to take questions. 
Vince, are there any questions for me?

00:18:07.340 --> 00:18:14.160
>> Yeah, they'll let me—let me get off 
of, uh, mute. There, yeah, that was great.  

00:18:14.880 --> 00:18:21.180
So, folks, this gives you an opportunity to ask 
all of your questions, um, to Ms. Chisnell. That  

00:18:21.180 --> 00:18:24.600
was a great presentation. I think that was a 
lot of information. I was actually making notes,  

00:18:24.600 --> 00:18:30.180
like always. I was trying to catch them the 
last part, "Equitable, accessible," and then  

00:18:30.180 --> 00:18:34.560
I lost you. You took the screen down right when 
I was trying to get that last part written down.

00:18:34.560 --> 00:18:44.820
>> I'll bring it back, okay? Could I do this again 
twice without—oh, you're doing great. All right,  

00:18:46.200 --> 00:18:50.760
that's about how to do this, is what I want 
to do. There you go. Equitable, accessible,  

00:18:50.760 --> 00:18:54.740
and human-centered services, we build trust.

00:18:54.740 --> 00:19:05.760
>> Human-centered services, this is great. 
Hopefully, everybody on the call really enjoyed  

00:19:05.760 --> 00:19:11.100
that because I really did. Um, this is some 
good stuff. I was, um, I always like to take  

00:19:11.100 --> 00:19:15.300
a lot of notes from all the presenters, and I 
really like the way your presentation kind of  

00:19:15.300 --> 00:19:21.960
broke it down in really big chunks, but it wasn't 
daunting chunks of information, but it made sense,  

00:19:21.960 --> 00:19:26.400
right? So, I was just kind of putting down 
some notes here, like the inclusive design,  

00:19:27.300 --> 00:19:32.040
accessible plain language, cultural 
change, partnership and engagement,  

00:19:32.040 --> 00:19:38.040
modernization technology services, especially 
the last part about the non-citizen informed  

00:19:38.040 --> 00:19:43.260
compliance for the multi-lateral information for 
those who are not speaking the same language.  

00:19:44.280 --> 00:19:50.280
This is just a lot of great, great information. 
And now, people, of course, are saying, "Oh,  

00:19:50.280 --> 00:19:55.080
thank you." They want to get access to, 
you know, your presentation. But I really  

00:19:55.080 --> 00:20:00.240
would like for you guys to ask any questions 
that you can think of. So, here's—okay, now  

00:20:00.240 --> 00:20:05.280
they're starting to come in. Thank you so much. 
Let's see if I can find one here for you, Dana.

00:20:05.280 --> 00:20:08.760
>> Looks like Gary has a question, Gary Morin.

00:20:08.760 --> 00:20:15.240
>> Okay, so is the focus on migrants in 
non— and let's see what happened here.  

00:20:16.320 --> 00:20:20.340
It looks like it broke apart. 
Or over U.S. disabled persons.

00:20:20.340 --> 00:20:27.180
>> So, oh, so it's a good question, 
and I can see why you asked that.  

00:20:27.720 --> 00:20:33.180
Um, I'm gonna stop sharing. 
Um, that's all right. Uh,  

00:20:33.180 --> 00:20:39.960
we, uh, I use these examples because they're very 
present in the work that we're doing right now,  

00:20:39.960 --> 00:20:50.700
but our focus is both on the employee experience, 
like what do you need to be the highest performing  

00:20:50.700 --> 00:20:58.260
person you can be in your job, we want to 
make sure that you have what you need to  

00:20:58.260 --> 00:21:05.400
meet the mission. And so, one of the things that 
we talk about is, uh, Tool Time versus goal time.  

00:21:07.260 --> 00:21:16.980
If you're focused on, uh, the technology that 
you're using, uh, that the department is supplying  

00:21:16.980 --> 00:21:25.080
for your job, uh, then that means that you're 
not focused on delivering on the mission. So,  

00:21:25.080 --> 00:21:36.480
for example, uh, I watched an, uh, ICE officer 
processing someone, and they had to copy and paste  

00:21:36.480 --> 00:21:47.400
information from five different systems into one 
PDF, and then print that out, get a wet signature  

00:21:47.400 --> 00:21:56.340
from the person they were processing, and then 
re-scan that document into a system. That's very  

00:21:56.340 --> 00:22:05.100
time-consuming. This is not what the age—like, 
these are highly qualified, highly trained agents,  

00:22:06.000 --> 00:22:14.880
uh, and they're spending all this time and energy, 
uh, on this burdensome administrative stuff.  

00:22:16.140 --> 00:22:26.520
Add to that, uh, having to do this with 
maybe sub-optimal tools that are not, uh,  

00:22:26.520 --> 00:22:35.280
accessible to them because they have some level of 
disability, uh, and now it just becomes a major,  

00:22:35.280 --> 00:22:42.720
uh, a major ordeal. And looking at this 
from the person they're processing, uh,  

00:22:42.720 --> 00:22:51.420
from the point of view, there, if that person 
comes in with medical issues or disabilities, uh,  

00:22:51.420 --> 00:23:02.100
it is often the case that we hand people a printed 
document in English, uh, to show what we have done  

00:23:02.100 --> 00:23:13.740
in this interaction. And, uh, they may or may 
not have, uh, the abilities or the skills to  

00:23:13.740 --> 00:23:20.520
understand what that is, which disadvantages 
them in whatever process they're part of.  

00:23:22.800 --> 00:23:27.120
But the truth is, every day, we're working 
with all different parts of the department,  

00:23:28.260 --> 00:23:36.120
um, on employee-facing tools and 
processes, as well as public-facing, uh,  

00:23:36.120 --> 00:23:43.140
services that are digital and analog, 
so it really runs the full range.

00:23:43.140 --> 00:23:51.180
>> Excellent. Excellent. Someone said inordinate 
experience impact outward experiences as well.

00:23:51.180 --> 00:23:52.500
>> Love it.

00:23:53.300 --> 00:23:56.280
>> All right, any other questions, folks?  

00:24:02.580 --> 00:24:08.520
Looks like, I said, okay, how do we, how do we 
serve and support non-Americans with disabilities,  

00:24:08.520 --> 00:24:14.640
foreign sign languages, print or digital 
materials in languages other than English,  

00:24:14.640 --> 00:24:18.300
etc., such as CISA? Kind of what you were 
talking about there a little bit but—

00:24:18.300 --> 00:24:28.020
>> Yeah, this is, this is a huge challenge. Uh, 
CRCL recently updated the Department language  

00:24:28.020 --> 00:24:37.020
access plan, but there's a lot of work to do 
every day to operationalize these things. So,  

00:24:37.020 --> 00:24:49.920
uh, one thing that we can do for, um, for making 
information more accessible in other languages is,  

00:24:49.920 --> 00:24:57.600
uh, use video and audio. Uh, we could have audio 
readouts of printed documents. This helps lots  

00:24:57.600 --> 00:25:08.640
of other people too, right? Uh, and, uh, or, uh, 
we can use animated video that has a voiceover,  

00:25:08.640 --> 00:25:18.000
so you could make it available in lots of other 
languages and in lots of other printed languages  

00:25:19.020 --> 00:25:25.920
without a major lift. The key is, like, we're 
never going to—okay, I shouldn't say never.  

00:25:25.920 --> 00:25:37.200
It will be challenging to meet a goal where we 
can communicate with everyone in the language  

00:25:37.200 --> 00:25:49.320
that they have, but we can prioritize a large list 
and expand our offerings across the department.  

00:25:51.120 --> 00:25:57.720
Different programs in DHS serve different 
levels of language access right now,  

00:25:59.460 --> 00:26:06.120
and some of that is because they have done 
the work to know who their audience is,  

00:26:07.800 --> 00:26:17.520
and some of it is prioritizing the most 
frequently encountered languages and  

00:26:18.060 --> 00:26:28.200
then working out to folks who, um, have 
languages that are, that we see less so.

00:26:29.160 --> 00:26:32.400
Yeah, you just made me think of a couple 
of things. I'm a geek at heart, of course,  

00:26:32.400 --> 00:26:36.240
so I was just sitting there thinking to myself 
as you were talking about, like, how do we  

00:26:36.240 --> 00:26:41.880
provide it through some form of multimedia, and 
then the first thing came to mind to me was AI,  

00:26:41.880 --> 00:26:47.760
using some form of Bot where it doesn't have to be 
an individual person, right? But then the person  

00:26:47.760 --> 00:26:53.460
could put in, "This is my native language that 
I want to be able to convert over from English  

00:26:53.460 --> 00:26:58.200
so I really understand exactly what is being 
asked of me." So I did make a note of that.

00:26:58.200 --> 00:27:06.000
>> So, yeah, so a few years ago, uh, when I first 
came back to Federal service in the spring of,  

00:27:06.000 --> 00:27:12.060
uh, 2021, I had done a study where we 
were looking at immigration and using  

00:27:12.060 --> 00:27:19.560
a dedicated device for translation, um, 
in places where it was difficult to get a  

00:27:19.560 --> 00:27:26.820
human interpreter in the interaction 
quickly, and, uh, two years later,  

00:27:29.160 --> 00:27:37.200
uh, agents or and officers are regularly using 
Google translate, yep, on your device. It's on the  

00:27:37.200 --> 00:27:46.380
most people who they interact with, uh, also have 
devices that have Google translate on them, and,  

00:27:47.160 --> 00:27:57.120
um, while there are certainly imperfections, 
uh, it uh has been reasonably effective  

00:27:57.120 --> 00:28:04.620
in helping in many interactions where 
we're just not gonna find somebody who's  

00:28:05.220 --> 00:28:12.920
speaking, uh, an indigenous Guatemalan 
language, uh, you know, in a Port of Entry.

00:28:12.920 --> 00:28:16.260
>> Exactly, excellent.  

00:28:16.920 --> 00:28:21.840
All right, do we have any more questions? This 
is—I'm enjoying this conversation, by the way.

00:28:21.840 --> 00:28:27.360
>> Oh, good, thank you. I could talk about 
plain language and language access all day long.

00:28:27.360 --> 00:28:35.280
>> Oh, but you'll see, uh, okay, somebody 
said, "Oh, it was Nic." Nic came out and  

00:28:35.280 --> 00:28:42.240
actually jumped in and said, AI-powered earbuds 
for auto language translation, I like it.

00:28:42.240 --> 00:28:44.340
>> A universal translator, oh my God.

00:28:44.340 --> 00:28:49.200
>> Yeah, it's in the future, 
we're going to get there.

00:28:49.200 --> 00:28:51.360
>> No, no, it exists now, there's a company  

00:28:51.360 --> 00:28:56.940
selling one that can translate to 132 
languages, and, uh, the Google buds,  

00:28:56.940 --> 00:29:01.080
they came out and one of their selling 
features was the auto translation piece.

00:29:01.080 --> 00:29:04.740
>> Okay, so you know, since nobody 
else has the question, you know,  

00:29:04.740 --> 00:29:08.820
I want to know a little bit more about 
these, says you brought it up, Nic.

00:29:08.820 --> 00:29:10.197
>> Yeah, would you mind, Nicshan?

00:29:10.197 --> 00:29:10.800
>> How does it work?

00:29:11.840 --> 00:29:22.080
>> Yeah, so, you know, everyone's on, on, uh, what 
is it, those streaming, um, my wife worked for  

00:29:22.080 --> 00:29:31.260
them, I should remember this, people, um, they 
talk, right? Yeah, so, TikTok largely uses AI  

00:29:31.260 --> 00:29:37.740
to auto-translate from various languages to make 
people's videos available on different platforms.  

00:29:38.520 --> 00:29:44.280
So that same thing, there are several companies 
right now that focus on AI-generated language  

00:29:44.280 --> 00:29:49.980
and auto-translation. So there are a couple 
of companies that also come out with earbuds,  

00:29:49.980 --> 00:29:53.040
and if we know anything, if we've been 
following the earbud market lately,  

00:29:54.120 --> 00:29:59.760
um, they're being used for three purposes: 
one is for music, a second one is to work  

00:29:59.760 --> 00:30:06.060
as non-medical device hearing aids, and 
the third being for auto-translation.  

00:30:06.720 --> 00:30:12.900
So you'll start to see that a lot coming out to 
the market is enhanced features but also, too,  

00:30:13.740 --> 00:30:19.200
most of you are carrying smart devices, 
you probably already have two to three  

00:30:19.200 --> 00:30:22.500
language translators built into your smart 
device depending on which one you're using.  

00:30:22.500 --> 00:30:29.280
Microsoft has its own auto-translation, Google 
has its own auto-translation, and then Apple has  

00:30:29.280 --> 00:30:33.960
its own auto-translation as well, and those 
things are often, uh, they're using, we're  

00:30:33.960 --> 00:30:41.120
sharing machine learning, so they're improving 
by leaps and bounds almost monthly, it seems.

00:30:41.120 --> 00:30:49.740
>> Interesting, that was awesome, awesome. 
All right, let me see if there's a recording,  

00:30:51.420 --> 00:31:00.180
see if anybody, uh, is adding any other comments 
regarding the Google Translate example the Dana  

00:31:00.180 --> 00:31:06.060
shared. Are folks on the ground capturing those 
repetitive exchanges to help with informed  

00:31:06.060 --> 00:31:11.940
content planning? So many opportunities 
are there. This is from Miss Rodriguez.

00:31:12.800 --> 00:31:17.820
>> That is a great idea. I don't 
know the answer to that question,  

00:31:17.820 --> 00:31:22.140
but I'm going to go back and find 
out because we should be doing that.

00:31:22.140 --> 00:31:28.260
>> Yeah, excellent, yeah, 
thumbs up on that one, Karla.  

00:31:29.820 --> 00:31:34.860
Um, okay, how to go about clearing 
the technologies in classified spaces?

00:31:34.860 --> 00:31:43.980
>> Ah, oh, this is really hard, is 
this from Alex who asked this question?

00:31:43.980 --> 00:31:47.700
>> It is just from a guest, 
they didn't say who they were.

00:31:47.700 --> 00:31:58.020
>> Okay, uh, yeah, this is a real challenge. 
Uh, I know that, uh, first of all, uh,  

00:31:58.740 --> 00:32:06.900
you know, SCIFs are not outfitted, and, um, and 
also, some of the equipment that you could take  

00:32:06.900 --> 00:32:16.440
in there that is accessible is very old, uh, and 
so, uh, we have a lot of work to do on that. If  

00:32:16.440 --> 00:32:22.200
there are folks in this audience who would like 
to work on that, um, please send me a note,  

00:32:22.200 --> 00:32:28.800
and let's figure out how to organize that because 
this is not getting to be a smaller problem,  

00:32:29.700 --> 00:32:37.800
uh, over time, and, um, yeah, I'm not quite 
sure how to attack it, so very open to ideas.

00:32:37.800 --> 00:32:42.480
>> So, that, that question actually 
came from Karla Rodriguez. She said,  

00:32:42.480 --> 00:32:44.700
"Awesome, would be happy to help."

00:32:44.700 --> 00:32:47.100
>> Lovely, thank you so much.

00:32:47.100 --> 00:32:54.840
>> And then, uh, let's see, Owen couldn't 
ChatGPT used as a language translator for those  

00:32:54.840 --> 00:32:58.980
in the field, probably. So, I think that's 
kind of early in the stages for a ChatGPT,  

00:32:59.580 --> 00:33:03.240
which is less than a year old, 
but I'm sure it could be, Owen.

00:33:03.240 --> 00:33:11.700
>> Yeah, there are, there are versions out 
there or APIs that use the engine that do  

00:33:11.700 --> 00:33:19.980
do translation. There's a whole lot to say about 
large language models generally, like these are,  

00:33:21.480 --> 00:33:31.380
uh, basically auto-complete tools, um, 
that are not smart themselves. Right,  

00:33:31.380 --> 00:33:34.860
they, they're content generators, 
they are not truth tellers.  

00:33:36.180 --> 00:33:42.240
Um, and we think that they're not truth tellers in 
lots of ways, right, like they do make things up,  

00:33:43.560 --> 00:33:51.420
um, because they're extrapolating on their 
algorithm, which is the thing that tells them,  

00:33:51.420 --> 00:33:59.520
uh, to mimic human conversations. 
So, uh, there's some interesting  

00:34:01.200 --> 00:34:09.960
ideas that could come out of whatever is next from 
ChatGPT, but there are a lot of, uh, as someone  

00:34:09.960 --> 00:34:20.280
pointed out in the chat, security concerns and 
issues, uh, containing it from what it is, uh,  

00:34:20.280 --> 00:34:29.340
actually, um, learning from and sourcing, uh, 
so you probably get to see experiments over the  

00:34:29.340 --> 00:34:35.160
next few years, less around translation, and 
more around, like, customer assistance. So,  

00:34:36.000 --> 00:34:41.400
um, like virtual assistants will start, 
um, incorporating large language models  

00:34:41.400 --> 00:34:46.020
that are trained specifically 
on a closed set of information,  

00:34:47.400 --> 00:34:54.240
which could work, but might still make stuff 
up. I mean, I don't know how you prevent that,  

00:34:54.240 --> 00:35:04.140
and so there's, there's a lot of work to do there, 
um, but the Secretary created an AI task force,  

00:35:04.140 --> 00:35:10.440
uh, and, uh, these kinds of things are, 
uh, some of what they'll be looking at.

00:35:10.440 --> 00:35:14.700
>> Excellent. So, we have one more 
question from DaShunda McDonley.  

00:35:15.900 --> 00:35:22.080
Um, how's the department going about collecting 
internal customer experiences to inform CX goals?

00:35:22.080 --> 00:35:29.100
>> Yeah, uh, we're doing a few things. We're 
leaning heavily on human-centered design methods,  

00:35:29.100 --> 00:35:38.220
so, uh, you probably will see surveys and 
questionnaires around, but we are very  

00:35:38.220 --> 00:35:44.700
keen on learning much more about the lived 
experience through qualitative research. So,  

00:35:44.700 --> 00:35:53.880
you may be invited to, uh, to an interview that's 
confidential between you and somebody on my team,  

00:35:53.880 --> 00:36:02.160
uh, or, uh, maybe a small group of trusted folks 
who you're open to having a conversation with,  

00:36:02.160 --> 00:36:10.860
uh, or, uh, opening up other lines of, uh, 
communication. But what we really love to do  

00:36:10.860 --> 00:36:16.740
is field research, where we can go to where you 
are doing the work that you are doing, so we can  

00:36:16.740 --> 00:36:23.880
see what it's like, uh, every day. So, if you're 
disabled, for example, and you are working from  

00:36:23.880 --> 00:36:33.540
your home office, uh, and, uh, you, we're working 
on an employee experience that, that you have,  

00:36:35.040 --> 00:36:46.980
uh, that you work through every day, we might 
just, like, hang out with you for a few hours and,  

00:36:46.980 --> 00:36:53.220
um, so you can help us know what that experience 
is like. I know for my own personal experience,  

00:36:53.760 --> 00:36:59.520
um, before I joined the federal 
government, I had the chance to work  

00:37:00.300 --> 00:37:10.860
on accessibility for a large, uh, financial 
company, uh, whose name will go unspoken, uh,  

00:37:10.860 --> 00:37:15.660
that, uh, was really interested in making 
sure that their new website was accessible  

00:37:15.660 --> 00:37:23.520
to people with disabilities, and so we went to 
folks' homes, and, um, the first blind person  

00:37:23.520 --> 00:37:31.200
who we had as a participant in our study, 
uh, invited us in, it was really great,  

00:37:31.200 --> 00:37:35.820
but it was dark in the house and all the 
lights were turned off, and we sighted people,  

00:37:35.820 --> 00:37:42.960
and I had to ask to have the voice turned on so 
we, uh, so we didn't trip over something, and,  

00:37:43.500 --> 00:37:48.120
um, I think we all learned a lot that day, 
and, like, there's just no substitute for,  

00:37:49.140 --> 00:37:58.200
um, actually getting to, um, see what the work 
is like, see what the context is like, uh, and,  

00:37:58.860 --> 00:38:03.660
um, live a little bit of that 
lived experience for a snapshot.

00:38:03.660 --> 00:38:10.440
>> 100 percent. So, folks, we 
are actually at the 2:50 PM mark.  

00:38:11.580 --> 00:38:16.380
It's just now, Dana, thank you so much, 
this was great, thank you for being, uh,  

00:38:16.380 --> 00:38:21.360
very candid and open and just making this more 
conversational, uh, for myself and everybody  

00:38:21.360 --> 00:38:26.160
that's on the call. I think it does us great 
when we have these opportunities to have these  

00:38:26.160 --> 00:38:30.840
type of conversations and just let ideas flow, 
because I think that's what it's all about,  

00:38:30.840 --> 00:38:36.420
having these collaborative moments. If there are 
not any other questions, we are going to close  

00:38:36.420 --> 00:38:42.240
out for today with closing remarks by Mr. Nicshan 
Floyd. Again, Miss Chisnell, thank you so much.

00:38:42.240 --> 00:38:43.680
>> Great, thank you so much.

00:38:43.680 --> 00:38:53.520
>> All right, so folks, we are going 
to close out, um, our day one of