00:00:00
when I started doing spatial analytics
00:00:01
full time I really struggled to build
00:00:03
out my skills and toolkit it took me
00:00:05
three years to land my first job at
00:00:07
cardo and about four years after that I
00:00:09
actually transitioned to a full-time
00:00:11
technical role during that time I tried
00:00:13
lots of different ways to build my
00:00:15
technical skills some of those worked
00:00:17
and some of those didn't I spent tons of
00:00:19
time reading medium articles taking
00:00:21
courses on udemy and tapping on
00:00:23
colleague's shoulders to ask annoying
00:00:25
questions on how to do a really simple
00:00:27
task so I asked myself if I had to
00:00:29
completely start over by learning
00:00:31
technical or modern GIS what steps would
00:00:34
I take to get there as fast as possible
00:00:36
so the question is can you move into
00:00:38
modern GIS or spatial analytics faster
00:00:40
100 this is exactly what this video is
00:00:43
about removing all the fluff and
00:00:45
focusing on the key steps that you need
00:00:46
to take to learn technical geospatial
00:00:49
analytics I get multiple messages on
00:00:51
LinkedIn asking about my exact process
00:00:53
or tips to do this so I figured why not
00:00:55
make a video about it stay tuned for the
00:00:57
whole video because I'm going to share
00:00:58
my top three mistakes that I myself and
00:01:00
I see others making when learning more
00:01:02
technical or modern GIS so the first
00:01:05
thing you need to do is pick your tool
00:01:06
set what tools should I use and in which
00:01:09
order should I learn them you're going
00:01:10
to need to have a combination of tools
00:01:11
together some things to visualize data
00:01:13
other places to store your data and then
00:01:16
some languages to analyze that data so
00:01:18
what's the first thing to get started
00:01:19
with my recommendation is getting
00:01:21
started with qgis this is still the best
00:01:24
way to get started it's free to download
00:01:25
and anyone can use it plus there's an
00:01:27
amazing community of support around this
00:01:29
tool is this still the tool that I use
00:01:30
the most today no would I change
00:01:32
anything about the order I learned this
00:01:34
in no way you can do everything from
00:01:36
analyzing spatial relationships reading
00:01:38
and analyzing raster files even up to
00:01:41
simple spatial statistical models within
00:01:43
qgis and if you can't find something
00:01:45
that you need to do I almost guarantee
00:01:47
there's a plugin to do this if you're
00:01:49
coming from a more traditional GIS set
00:01:50
of tools this is a great place to start
00:01:52
it'll feel super familiar and it's very
00:01:54
easy to use there are of course some
00:01:56
limitations with qgis you can't analyze
00:01:58
super large data sets otherwise you
00:02:00
might end up with a spinning wheel of
00:02:02
death like you see here
00:02:07
once you've learned this what's the next
00:02:09
thing that you need to learn so once
00:02:11
you've learned qgis it's time to move on
00:02:13
to your next tool now you're going to
00:02:14
need a programming language to do more
00:02:16
spatial analytics and analyze larger
00:02:18
data set previously I said that python
00:02:20
was the best way to get started with
00:02:21
this and I'm actually going back and
00:02:23
changing that recommendation a little
00:02:24
bit the next tool that I would learn is
00:02:26
spatial SQL so I SQL over python the
00:02:29
answer is really simple the faster your
00:02:30
ability to query change and manipulate
00:02:33
data the faster your career will move as
00:02:35
a spatial analyst for me this is one of
00:02:37
the things that catapulted my career
00:02:38
forward and so that's why I'm
00:02:40
recommending it here because I feel like
00:02:42
as a next logical step this is the way
00:02:44
to go spatial SQL is highly popular and
00:02:46
really in demand right now especially
00:02:48
for lots of different roles and can be
00:02:49
helpful across the board but spatial SQL
00:02:51
is really hard to learn even in
00:02:53
traditional settings I had to learn this
00:02:55
by picking up random tutorials asking
00:02:57
colleagues and kind of figuring it out
00:02:58
as I go along but like I said one of the
00:03:00
most valuable skills that I've learned
00:03:02
to date and I say that for two different
00:03:04
reasons the first is that these
00:03:06
functional data skills that you're going
00:03:07
to build in SQL are highly transferable
00:03:09
inside and outside of Geo the second is
00:03:11
that this really just helps you scale up
00:03:13
what you're already doing this is also
00:03:15
going to help you build really practical
00:03:16
data skills doing things like ETL and
00:03:18
transforming your data and it's also
00:03:20
going to help you think about data in a
00:03:21
new way programmatically there's also a
00:03:24
pretty low barrier to entry SQL is
00:03:25
pretty easy to learn once you have a few
00:03:27
foundational tools to do so and you can
00:03:30
use it as a stepping stone to different
00:03:31
data types and other languages I've been
00:03:33
building a series all about spatial SQL
00:03:35
you can check that out in this playlist
00:03:36
here the best part is you can get
00:03:38
started at no cost you can download post
00:03:40
GIS connect it to qgis and you're up and
00:03:43
running in a few minutes now of course
00:03:45
you can analyze data to the blue in the
00:03:47
face ultimately you're going to have to
00:03:48
create a map and visualize your data so
00:03:51
what should your visualization toolkit
00:03:52
look like so there's no right answer
00:03:54
here and there's no clear consensus on
00:03:56
what tools you should use to visualize
00:03:58
your data a lot of times you're going to
00:03:59
see in job postings people asking for
00:04:00
skills in Tableau or power bi but
00:04:03
business intelligence tools just aren't
00:04:04
suited for geospatial data they can't
00:04:06
handle the volume of data and once you
00:04:08
get it above a certain number of
00:04:09
features they're going to crash so what
00:04:11
are the tools that you should learn if
00:04:13
you're focusing specifically on
00:04:14
geospatial data well the first again is
00:04:16
qgis qgis lets you visualize your data
00:04:19
within the application itself you can
00:04:21
create map exports that are static or
00:04:23
print versions and you can even create
00:04:25
some lightweight interactive maps as
00:04:27
well it's completely free and open
00:04:28
source and if you're a GIS team this is
00:04:30
going to be a really great way to get
00:04:31
started and if you're thinking about
00:04:32
Enterprise GIS if you're all connecting
00:04:34
to the same database you can all share
00:04:36
data back and forth and use that as your
00:04:38
core data store for your team that said
00:04:40
there are some limitations it's hard to
00:04:42
share maps and data back and forth you
00:04:44
have to transfer your file back from one
00:04:45
to another so sharing becomes a little
00:04:47
bit more tedious when you're using qgis
00:04:49
the other tool I really like is called
00:04:51
Kepler GL this uses deck GL as its
00:04:54
rendering Library which is really great
00:04:55
for the modern web it has lots of
00:04:57
comprehensive data visualization methods
00:04:59
and even has some different controls to
00:05:01
filter your data and then ultimately
00:05:02
share and publish maps all that said you
00:05:05
do have to self-publish your own maps
00:05:07
and you also can't connect a data source
00:05:10
like a database or something like that
00:05:11
so you're going to have to take your
00:05:12
data out and put it back into Kepler so
00:05:14
those are the core limitations there the
00:05:16
other options using cardo Now spoiler
00:05:18
alert I do work for cardo so I am a
00:05:20
little bit biased here but I do feel
00:05:21
like it offers some great options
00:05:23
compared to qgis and Kepler Carter lets
00:05:25
you connect any data source whether that
00:05:27
be a database or data warehouse
00:05:28
seamlessly and you can use that to build
00:05:30
Maps visualize them and share them it
00:05:32
also adds some different tools into the
00:05:34
database or data warehouse in its
00:05:36
analytics toolbox to make things like
00:05:37
spatial statistics routing geocoding and
00:05:41
creating map tiles even easier you can
00:05:43
create really complex dashboards with
00:05:45
writing SQL or without there are some
00:05:46
limitations cardo is completely
00:05:48
cloud-based and does have a cost
00:05:50
associated
00:05:51
if you do want to get started you can
00:05:52
get started for free with a trial or if
00:05:54
you're a student you get free access to
00:05:56
the GitHub student developer pack
00:05:57
there's no right answer here but in
00:05:59
terms of how you might start with these
00:06:00
tools I recommend starting with qgis to
00:06:03
do your base visualization in analytics
00:06:05
and then when you need to create an
00:06:06
interactive map moving that into Kepler
00:06:08
and taking your data out of qgis and
00:06:10
putting it into Kepler once you get more
00:06:12
proficient in SQL and you're ready to do
00:06:14
some more complex visualizations or
00:06:16
scale up with larger data I would take a
00:06:18
look at cardo because that becomes a
00:06:19
logical time to use a more scalable tool
00:06:21
so we have to add one more tool to our
00:06:23
toolkit and you might know what I'm
00:06:24
going to say
00:06:28
but it's python when you want to get
00:06:31
into more advanced analytics you really
00:06:33
need to add another programming language
00:06:34
to do this and python is the best choice
00:06:36
to do so my first programming language
00:06:38
was actually JavaScript and while this
00:06:40
taught me a lot and I had to struggle to
00:06:41
learn it I wouldn't recommend that as
00:06:43
your analytical programming language
00:06:45
python is going to give you a great base
00:06:46
to work off of and scale your skills
00:06:48
well beyond this language itself I also
00:06:50
get this question all the time should I
00:06:52
learn r or should I learn python
00:06:54
ultimately I would recommend python R is
00:06:56
really used in academic circles and it's
00:06:58
a really great toolkit to get started
00:07:00
but python not only as an analytical
00:07:02
language helps you do things like
00:07:04
process data create data engineering
00:07:06
pipelines run in data science notebooks
00:07:09
create machine learning models Even
00:07:11
build back-end apis it's extremely
00:07:13
versatile so once you learn it you can
00:07:14
apply to much more things Beyond just
00:07:16
spatial analytics python is super easy
00:07:18
to get started there's a lot of courses
00:07:20
to learn but you can take a look at this
00:07:22
video which has my recommendations on
00:07:24
how to get started and Learn geospatial
00:07:25
Python so now that we know what to learn
00:07:27
we have to figure out how to learn it
00:07:29
and this is the first mistake that I see
00:07:31
a lot of people make people try to learn
00:07:32
everything that they can possibly learn
00:07:34
with all these different languages and
00:07:36
tools and this isn't the approach I
00:07:37
would recommend you want to become just
00:07:39
dangerous enough to use these different
00:07:40
skills in practical ways and then over
00:07:43
time as you pick your focus areas you
00:07:45
can really start to go deep if that's in
00:07:46
spatial data science with python or in
00:07:49
data engineering or SQL you can really
00:07:51
figure out where your interests lie the
00:07:52
other problem with this is that people
00:07:54
spend a lot of time watching tutorials
00:07:56
and I mean really watching them the
00:07:58
number one thing that you can do is
00:07:59
actually practice writing code even if
00:08:02
this is a simple hello world statement
00:08:04
or building your skills over time if
00:08:06
you're practicing writing code you're
00:08:07
building your skills courses are really
00:08:09
great and I've learned a lot from them
00:08:10
over time but there's a lot of great
00:08:12
free tools out there that help you get
00:08:14
started and actually have practical
00:08:15
exercise to do this I shared this
00:08:17
earlier in my videos for geospatial
00:08:19
Python and the first video on my spatial
00:08:21
SQL course as well the other thing you
00:08:23
need to know is analyzing geospatial
00:08:24
data in practice is way different than
00:08:27
watching it so getting your hands dirty
00:08:28
hitting some walls and trying to figure
00:08:30
out problems is the best way to go
00:08:32
knowing how and what a spatial joint is
00:08:34
is way different when you have to join a
00:08:36
couple million points to a couple
00:08:37
hundred thousand polygons what's the
00:08:39
best way to practice there's a few
00:08:40
things that I can recommend the first is
00:08:42
to design your own challenges you can
00:08:44
actually think of here's a problem I
00:08:45
want to solve and how would I solve that
00:08:47
with python or SQL and decide how you
00:08:50
want to go from there there's so many
00:08:51
great open data sets to test this out
00:08:53
basically every city or country in the
00:08:55
world has an open data portal so you can
00:08:57
go and grab some data and start solving
00:08:58
different problems that way you can also
00:09:00
look at different data sources like
00:09:01
Google open data as well as there's lots
00:09:03
of geospatial data sets on things like
00:09:05
kaggle and other places too another
00:09:07
great way I recommend is looking on
00:09:08
medium and trying to find projects that
00:09:10
you like and find interesting and
00:09:11
replicate them or add a different
00:09:13
spatial angle to them find three maybe
00:09:15
four projects that you really find
00:09:17
interesting and take your own spin at
00:09:19
them from a geospatial perspective I'd
00:09:20
also recommend investing maybe in a few
00:09:22
different training tools that are
00:09:24
specifically focused on helping you
00:09:25
practice there's a few that I really
00:09:27
like strata scratch which is built by
00:09:28
data scientists and actually puts you
00:09:30
into programming challenges for SQL as
00:09:32
well as python if you're focusing purely
00:09:34
on SQL learnsql.com is another awesome
00:09:36
resource that I really like there's lots
00:09:38
of tutorials you can take everything or
00:09:40
just the bits that you want to focus on
00:09:42
the last one is data by Danny if you're
00:09:43
going deep deep into SQL this is the
00:09:46
best route to go there's an eight week
00:09:48
sequel challenge that he has as a
00:09:50
complete course that really focuses on
00:09:52
deep and intense topics you can do all
00:09:54
of this totally for free with no cost
00:09:56
but if you're going to invest in a tool
00:09:57
I would definitely invest in one that
00:09:59
gives you actual problems to work on and
00:10:01
train the second biggest problem I see
00:10:02
people make is feeling like they have to
00:10:04
figure it all out by themselves guess
00:10:06
what you don't in analytics or
00:10:08
geospatial analytics something is
00:10:10
inevitably going to go wrong no God
00:10:12
please no no no no and you're gonna have
00:10:16
to figure out how to fix it this is
00:10:17
where you're gonna spend a ton of your
00:10:18
time working in technology no matter
00:10:20
where you are you're inevitably going to
00:10:22
end up on stack Overflow at some point
00:10:23
trying to figure out how to do something
00:10:25
my other advice is to learn how to read
00:10:27
error codes and learn how to read
00:10:29
documentation error codes can be super
00:10:31
annoying but a lot of the times if you
00:10:32
can read them and understand what the
00:10:34
problem is that might give you a clue is
00:10:35
where two the problem might be and also
00:10:37
reading documentation to see what needs
00:10:39
to go into a function and what comes out
00:10:41
of it is going to be a valuable skill as
00:10:43
you start to get deeper and deeper into
00:10:44
these tools so now we've picked our
00:10:46
tools we've learned them and now it's
00:10:48
time to find a job and this is actually
00:10:49
where I see people make the third most
00:10:51
common mistake and that's not being
00:10:53
proactive what do I mean by being
00:10:55
proactive it's actually three things
00:10:56
first building your portfolio second
00:10:59
creating and building out your LinkedIn
00:11:01
profile and third reaching out to people
00:11:04
that you might want to work with so the
00:11:05
first step is building out a portfolio
00:11:07
why is this so important building a
00:11:09
portfolio shows that you first of all
00:11:11
know how to do the work but also that
00:11:14
you've actually had some practical
00:11:15
implications for this so how do you
00:11:16
start building portfolio projects if you
00:11:18
can actually do this in your current job
00:11:20
great that's a great place to start and
00:11:22
actually apply your skills and build
00:11:24
some different projects if not try to
00:11:26
find some projects or some passion
00:11:28
projects that you want to work on no
00:11:29
matter what the project you do make sure
00:11:31
it's Unique to you and you're passionate
00:11:33
about it that's going to shine through
00:11:35
no matter what that is if you're really
00:11:36
passionate about the outdoors do
00:11:38
something with national parks data and
00:11:40
try to figure out which Parks have the
00:11:41
most visitors are you passionate about
00:11:43
cities Great go find some open data and
00:11:45
build a project there another tip is you
00:11:47
can even reach out to non-profits or
00:11:49
businesses figure out if they have a
00:11:50
geospatial problem and actually help
00:11:52
solve it for them you can start this for
00:11:54
free or turn into a freelancing side
00:11:56
hustle as well I'll go into portfolio
00:11:58
projects in a future video in Far more
00:12:00
detail but just getting started is the
00:12:02
most important piece now where should
00:12:03
you put your portfolio frankly anywhere
00:12:05
building a simple website on WordPress
00:12:07
or even a simple HTML page is one way to
00:12:10
start make sure you get your work on
00:12:11
GitHub if you're hosting code this is a
00:12:13
great place to go and also check out
00:12:14
spatial node which is a portfolio tool
00:12:16
just for the geospatial community when
00:12:18
you present your portfolio projects they
00:12:19
should have three things they should be
00:12:21
short and to the point they don't need
00:12:22
to go into pages and you don't need to
00:12:24
have a 50 page report about them all
00:12:26
that said in the second point they
00:12:27
should have some detail in it talk about
00:12:29
what you did how you did it and what
00:12:31
were the outcomes that's actually the
00:12:33
third Point what are the outcomes and
00:12:35
what did you solve did you find a new
00:12:37
trend did you uncover something in the
00:12:38
data Focus your point on that outcomes
00:12:41
are the focus of all geospatial
00:12:43
analytics so make sure you put that
00:12:44
front and center now it's time to move
00:12:46
in and optimize your LinkedIn how do you
00:12:48
do this there's great resources on how
00:12:50
to optimize your LinkedIn to get a job I
00:12:52
actually really like this video that
00:12:53
tells you a lot more about creating a
00:12:55
really effective LinkedIn profile using
00:12:57
keywords and adjusting key components
00:12:59
your profile to help you find a great
00:13:01
job that's going to be a good fit for
00:13:02
you in general try to think of LinkedIn
00:13:04
as a search engine if you want to become
00:13:06
a geospatial analyst put that in your
00:13:08
headline in different parts of your
00:13:10
profile you want to be a geospatial data
00:13:12
engineer great focus on that and add the
00:13:14
relevant skills now in a perfect world
00:13:16
you build all this and recruiters start
00:13:18
reaching out to you but unfortunately
00:13:19
geospatial you can't always count on
00:13:21
that geospatial is still a niche and
00:13:24
lots of recruiters unless they're
00:13:25
working on a company that's really
00:13:26
focused only on geospatial don't always
00:13:28
know the right things to search or
00:13:30
search for it's up to you to first of
00:13:31
all look for roles that might be a good
00:13:33
fit sometimes in a data analyst profile
00:13:35
you might find a listing for someone who
00:13:37
wants to build maps are you looking at
00:13:38
data engineering you might actually
00:13:40
learn that they might be using gdal and
00:13:42
that's a great way to search for that
00:13:43
here's another post to actually search
00:13:44
for different terms and optimize
00:13:45
researching when you're looking through
00:13:47
job listings themselves all this is
00:13:49
great but the number one tip I have for
00:13:50
LinkedIn is being proactive that
00:13:52
sharing your ideas posting about things
00:13:55
you're learning posting your portfolio
00:13:57
projects and most importantly reaching
00:13:59
out geospatial has one of the most
00:14:01
active and engaged communities on
00:14:03
LinkedIn and other social platforms and
00:14:05
people are always learning from and
00:14:07
sharing with each other take advantage
00:14:09
of that jump into the conversation and
00:14:10
share interesting things that you see
00:14:12
you're working on or you're excited
00:14:14
about did you build a great new
00:14:15
portfolio project great share it did you
00:14:18
find a new code snippet that was really
00:14:19
helpful for you great share that with
00:14:21
everyone as well the other thing I would
00:14:22
say is don't be afraid to reach out to
00:14:24
others if you see someone that's working
00:14:26
in a team or a role that you're
00:14:27
particularly interested in reach out it
00:14:29
might not work out the first time but
00:14:31
they might have something in the future
00:14:32
or they might know someone who needs
00:14:34
need of a geospatial expert like
00:14:35
yourself give it time keep up the hard
00:14:37
work and eventually it will get there
00:14:38
now in a lot of other spaces you'll see
00:14:40
a ton of emphasis on the technical
00:14:42
interview or program interviews
00:14:43
geospatial interviews vary widely so
00:14:46
what you want to do is be able to ask
00:14:47
questions and any information that you
00:14:49
can get beforehand is going to be really
00:14:51
helpful for you when you jump into those
00:14:52
first interviews if you had a tough
00:14:54
interview or something didn't work out
00:14:55
great figure out what that is and try to
00:14:57
change it for next time and re-implement
00:14:59
it keep practicing keep trying and
00:15:01
eventually you'll translate that into
00:15:03
success in the real world geospatial
00:15:05
analytics is booming so now is the time
00:15:07
to jump in with two feet start learning
00:15:09
growing and building your career
