Working With Open Source Vendors

About

Makepath is a spatial data science firm that works with all types of organizations in industries ranging from finance, resource management through to retail. Their specialty is bringing open source to the challenges inside the organizations.

This podcast will be really useful to people who want to understand how their enterprise can benefit from working with an open source software vendor. Some of the topics that we discuss include:

– How Makepath helps clients to solve burning problems with open source, in addition to connecting them directly with the open source communities their involved in

– And what makes the Geospatial open source community unique and where it is heading in the future

Transcript

hey everybody welcome back to the 17th
episode of open source for business
brought to you by open teams my name is
henry badri and today i talked with
pablo fuentes
and brendan collins who are both
founders of make path
make path is a spatial data science firm
that works with companies from all
different types of industries ranging
from
resource management finance and all the
way through to
retail their specialty is bringing open
source to the challenges inside
organizations
i think that this episode will be really
useful to enterprises who want to know
how they can actually benefit from
working with open source software
vendors or experts
some of the topics that we discuss in
this podcast include
how make path helps clients to solve
burning problems with open source
software in addition to exploring the
ways in which make path and open source
vendor can help
their clients to connect with open
source communities that they’re involved
in
and we also discuss what actually makes
the geospatial open source community
unique
and where it’s heading in the future
whether you are a user
developer manager or if you’re just
curious about the industry
open teams is the place to find
information news
training and support you need to thrive
with open source
now that the introductions are out of
the way let’s dive right in
[Music]
[Applause]
pablo and brendan thank you for joining
me in the podcast today hey good to see
you
henry thanks so much for having us it’s
great to have you on i’ve really been
looking forward to chatting to you both
now i thought we’d start with you
brendan how did you get where you are
today
um yeah so i uh have a background in
geospatial technology
i got interested in geography in the mid
2000s working in a gis
lab specifically looking at
environmental science and trying to find
tools that would help me solve some of
the problems that i was looking at on
the environmental science side
as i got into geospatial technology
i saw the power of coding and automation
and saw that there was a lot of great
integration with python so my
love of python came from a need to solve
some problems that i couldn’t solve
uh brute force um i think the first
major issue that i ran into that
required coding was
unzipping 200 000 zip files and it was
finally something i couldn’t do by hand
and so
went to coding then um in
wanting to push further
i organized a group of people to help
address
client problems using open source tools
and then somewhere down the road pablo
fuentes and i
met up and realized that we would be a
great pair to
start make path and a year ago we
started make path
to bring open source to
client problems and also find mechanisms
to fund open source technologies
so i know you’ve got a bit of history
working with the ceo of open teams
uh can you tell us a little bit about
that journey at anaconda
yeah so i first printed out the guide to
numpy at kinka at kinko’s in
2007 and i saw the name travis oliphant
on the front of that
and um you know started trying to
understand what this numpy library was
and from there serendipity brought me to
austin
and also put me in contact with travis
through the local python community
i started working alongside travis at
continuum analytics which then became
anaconda i had a the great privilege of
actually
doing a hackathon with travis so travis
and i traveled once to washington dc to
do a hackathon
which was an absolute dream for me and
so you know i i really love the work
that travis does and i use the libraries
that he created every day
okay and that would have been great he’s
a great guy to work with and he’s just
so knowledgeable and
his passion for open source is
unequivocal i
have never met anyone like him he’s so
giving but
pablo uh i know you’ve got quite an
entrepreneurial background so how did
you what’s your journey from
entrepreneur into open source what does
that look like
sure absolutely so um i mean i’ve always
been interested in
in numbers and and data and i used to be
in the investment world before going to
grad school
um after uh i got my mba at stanford i
started a product company
that helped people get jobs using mobile
phones so i was the ceo and founder of a
product company for 10 years and after
i sold that company in
2019 about a year after that brenda and
i
got together and started thinking about
working together
and i have a passion for building
products i have a passion for
uh coming up with solutions and it’s
ever since i was a little kid i like
maps my my dad used to take me out but
he was a
biologist and a professor in the
university
and he would take me up in the in the
hills and we would read topo lines and
and so i always had an interest in in
maps and and
all of the things that have to do with
understanding uh
understanding the physical space and now
yeah it’s been great uh
the the expertise that brendan brings to
the table on the open source side
is is unparalleled and it’s amazing i
feel really lucky and blessed to
to work with him as a partner and a lot
of what i bring to the table is the ways
to think about product and also the ways
to build a business and
uh you know as we’ve grown our our
cu our team and our company we you know
we have over a dozen people now
uh working working on make path and
we’re getting some great
uh uh companies we have partnerships
with fortune 50 companies we also have
uh some great uh open source grants that
we’ve gotten from
from people to to work on open source
initiatives uh
i’ve been learning more and more about
the open source community it’s not
something that i i mean i had a
passing understanding of open source
tools through some of the tools that we
used to build a product
uh in my last company but certainly the
last year has been
a very uh quick and
and brisk education and and uh
in all things open source and it’s just
it’s just a great
community to be a part of i think the
the notion and the the idea
that you give things to the community
and the community keeps giving back to
you
we hire most of the people that we hire
from people that contribute to open
source projects that we’re a part of
uh and we also uh find a lot of our
clients
and a lot of our partners from other
people through i guess people that who
are also interested in open source and
one one of the key principles of our
company we have this running list of
key principles that brendan and i run
the company by we have six of them and i
have a little note here that i look at
every day
and one of them is give things away
serve your community
and the rest will take care of itself
and that’s not only on the software side
but also everybody at make path
gets one day a month that they get paid
to go work in their local community and
volunteer in the local community in some
capacity
uh and really help uh their community uh
you know
face to face and obviously as kovit uh
is is beginning beginning to to subside
and hopefully by the end of this year
we’ll be
uh you know back in some semblance of
what used to be normal
um people will be able to get back in
face to face with their communities but
that idea of
getting more through giving more
is something that’s been a really
amazing part of working with the open
source community
i think that’s definitely very inspiring
and something a lot more companies
should be involved in i know that
companies are starting to take
uh contributing back to open source
communities a lot more seriously
a lot of companies are employing open
source contributors like yourself
and i think i really appreciate the idea
that you’ve taken the open source
esource an idea of giving back to
community
and brought it to the physical world of
not just an online community but to your
local community i think that
would make such a huge difference and i
i’m interested because i’m not currently
in austin there was a huge community
effort to help with the
recent problems uh that came from the
snow wasn’t there wha
what did that look like and how did the
community come together in austin
um paul i can cover that that one um
so there’s certainly uh there is
a lot of there were a lot of problems
and i’m not gonna say that it
came together in in some beautiful way
uh it was nice to to be able to
communicate with neighbors in that sort
of time a lot of crisis crisis brings
people together
and there was certainly that but um
that was an unprecedented blizzard for
the
central texas and and it is
was a really difficult time but that is
um
you know continues as as people repair
and they repair the
the grid but it was it was a difficult
is a difficult week
okay my personal situation is my wife
and i were actually
we’re supposed to be out of town for
three days that weekend and we were
unable to come back to austin because of
our flights kept getting canceled so i
we had our neighbors
uh check check on our cat unfortunately
we never lost power and our dog was
being taken care of by somebody else so
everything was
fine um but i i kept on texting people
and asking them if they needed it
because
our place never lost power and so if we
had friends come to our place and take
showers we had
i i have a bunch of wood in the side of
the house that i offered to people
that didn’t have power for so it could
heat their homes and also read it
of all things the number one source for
information of what was going on was the
austin reddit forum
where people were sharing uh information
of what was going on and ways to help
each other
and that’s open sourcing solutions
because really frankly
i love the great state of texas but the
government didn’t exactly do a stellar
job at managing communications or
or managing supplies or anything like
that nor nor did the city so the number
one source that i had for information
that i
deemed to be directionally accurate was
the austin reddit much more so than
local news outlets so that’s
in some ways a way to open source uh and
and community drive the answer to
a you know a weather crisis
i definitely agree and i think as we’re
all becoming more digital it
definitely has helped particularly
during covert i think that
people have jumped online and come
together as online communities to
support each other
but now just to shift gears a little bit
i’d like to focus on make paths so
for those listening what does make path
do and what kind of services do you
offer
so uh make path is a spatial data
science firm
um we work with a variety of clients in
different industries
everything from uh natural resource
management
to finance to um uh
retail and we bring open source to the
to the challenges inside these
organizations
many times those um they look at us for
expertise in geospatial technology
but we have a breadth of expertise in
data science topics
that range from distributed computing to
algorithm
optimization to specific um geospatial
analysis and also visualization
we love beautiful applications and think
that it’s
it’s um it’s really part of the
communication
of the insight we’re trying to get to
insights
using information and using data so when
we can communicate those insights in a
beautiful way
we you know we think that that is
a worthy thing to focus on and focus
resources
towards we also um really focus
on on nurturing early open source
developers
so people that are trying to get into
the community get visibility into the
community
and so a big part about of make path is
um
being a um you know a nurturing place
for new engineers and new devs that want
to get into open source
and and need a shot so we have uh one is
is serving our outside clients and we’re
very passionate about that but then also
serving our team and serving the the
community to
develop our engineers and help them with
their careers as they go forward
i think that’s great and i think again
hiring open source developers makes such
a huge
difference just on the community and and
the ecosystem as a whole we think
that is one of our missions that open
teams and the travis is very passionate
about is making open source thrive
and we think really that can only happen
when you can join
open source and become a part of a
community with the expectation that you
can make a living
out of your passion or hobby because
historically travis oliphant he started
numpy and scipy and
he had six kids and for a decade was
just
it was difficult to be able to get by
because he’s devoting so much of his
time to open source so he had to be
innovative and all the other ways of
renting out his time and i think now
that
companies are starting to get more
involved like yours like the big tech
companies
they’re starting to employ open source
contributors i’m really excited to see
the next
phase of open source with this
sustainability in mind this
idea that we want to help people
actually make a living and build a
career around open source
we think there’s a long way to go and
i’m very very excited about that
one one thing i wanted to add to that uh
is also
it’s a virtual cycle right so there are
companies
and organizations that are paying us
because they
care about the tools that we’re
developing so we have x-ray spatial that
was a tool that was pioneered by brandon
here on this call
and that is one of our main projects
alongside data shader and bokeh that we
that we have core contributors to
and there are companies are coming to us
that
see the value that these projects have
to their initiatives and to the
community
and they are engaging us and paying us
money
to build these tools up and the virtual
cycle happens because then when people
come to
us so developers we we’re you know we’re
constantly hiring so we’ll do like this
small plug here we are hiring so if
you’re interested in working
in open source uh please do reach out to
us at combat.com and we’d love to talk
but a lot of times when people reach out
to us and we have many conversations per
week with people who are interested in
working with us
we tell them hey the number one way in
which you can do
impress us or get in touch with us to
for us to start aligning
with each other is for you to make
contributions to
some of the open source projects that we
work on and it doesn’t doesn’t mean you
have to do huge commits that are going
to change you know they or even add
meaningful features it can start with
something as simple as adding
documentation
obviously documentation is always
something that’s very very important
with code and with libraries because
regardless of how brilliant
something that you write is if somebody
else can’t understand it uh
it doesn’t you know it goes against the
ethos of what we’re doing with open
source
so the virtual cycle continues because
people
are able to build their careers because
they’re making contributions that are
available in a public forum and
their github profile becomes a resume
and then we end up hiring them or other
people in the community end up hiring
them
for their contributions and then that
means that
everybody benefits from this virtual
cycle
in the ecosystem and that’s something
again we’re really excited to be a part
of
i think that idea that you’re hiring the
open source contributors and then
allowing companies to come in and work
with you
rather than working directly with the
community is
revolutionary in itself and something
we’re big big believers in at open teams
so what does that look like so you’re
hiring open source contributors and then
you’re going out and your clients are
they directly involved with the
community or are you the middlemen
in between the enterprise and the
community
yeah um so it varies some of our clients
actually don’t know that much about open
source but have a
burning problem that they need to solve
and so they’re they’re less inspired by
open source
and more inspired by the solutions that
open source brings
then there are other clients that come
to us specifically because of our
involvement in those projects
and those are that’s really fun because
a lot of times that means that they’re
excited about contributing back to the
project
they’re interested in doing spin-off
projects so in some of our engagements
we’ll find an
a new niche for an open source project
that didn’t exist
and we can go ahead and try to fill that
niche
one of those is a library called map
shader which is very recent it’s about a
month old
and that fills the niche of connecting
http to data shader
so data shader is a you know a great
library that was started by peter wang
and really grown by uh jim bednar
and we wanted to hook that up to um ogc
which is the open geospatial consortium
map services so we take a um
uh open specification from ogc and we
combine it with an open tool like data
shader
and then we have something that serves
the gis community
even though they may not be familiar
with numpy or
das or numba they can still benefit from
those technologies that
might be have been developed in a
different domain
i think that’s a really great thing
that’s come from enterprises working
with open source communities
the idea that you can discover new
innovations new technologies
and you brought up peter wang who
started data shader
and i know that travis and peter and
everyone at anaconda
that entire company has been responsible
for creating some incredible projects
uh dusk spun out of anaconda data shader
as you said
and i think it’s that pi data ecosystem
that idea of innovation and the open
source
ethos that they embody allows them to
just
create create create so that leads me on
to my next question
what are some of the most common
problems that you at makepath help your
clients to solve
so one of the things that we talk about
at makebath always is visualizing what’s
hiding in your data
right so a lot of times we work with
clients i have very very large data sets
to where traditional tools and like
don’t have
are not able to give you the insight
that you need to understand that data
so a lot of what we do is using our open
source
tools and scientific python to
understand
uh and create visualizations in in
custom visualizations
and that can take many different shapes
that can be fully developed uh
you know front to back apps that you
know it’s like a
fully standalone uh project it could
also take the shape of features and i
don’t know brendan do you want to talk a
little bit more about that
yeah yeah sure so just to um to add on
to what pablo is saying
a lot of times the problems are framed
in the classic
three v’s of big data right of variety
volume and velocity so we have people
that come to us
with hey we have just we have a mess on
our hands and we need someone to come in
and help us with this variety or they
say hey it’s well structured but it’s
just too big for our machine and it’s a
volume problem
um or hey we we have this down but slow
and we need to we need to take this
operation
and scale it and maybe that’s the
velocity side most of the clients have
all three
so you know we’re touching in in
different spots
and we’re looking at um a lot of the
work that was done
by travis and uh to to help solve these
problems
and that work comes in the form of numpy
and scipy and number
and conda and the amazing tools that are
in the pi data ecosystem
that help us solve those problems that
many of our clients don’t know about
that was a great explanation of the
different problems that clients come to
you with
and the idea that you can easily sum up
those problems to the three v’s the
variety volume and velocity since you’ve
worked a lot with open source
communities over the years
at anaconda and now at make path i’d
like to ask what are some of the key
things that you’ve learned working with
these open source communities
and building these open source tools
that enterprises use
yeah i i think that um there’s
there’s a couple of guiding principles
one of them that we use at make path is
that we really want to put
the user first and think about that the
user problem
before we become too prescriptive with
technologies or tools
we we love the tools that we work on but
ultimately we’re about
change and being able to use the best
tools for the problem
so putting the problem first and making
sure that the open source tools that
that we’re creating keep that user in
mind and speak to that user
in the case of x-ray spatial we’re
speaking to the geography community
so we’re using language that they know
we’re meeting them where they are
as opposed to trying to um super you
know to
reinvent some nomenclature and that’s a
lot of
of what open source is about is like
what culture does this tool come from
and
and who does it serve exactly so in our
case in
xray spatial we want to have a function
called slope
because geography people understand what
slope is we don’t want to call it
gradient we don’t want to call it first
derivative
and so meeting people where they’re at
is a key part of open source
and then understanding that a lot of
people are dedicating
free time to these projects and um
there is a sometimes
neglect of the fact of how much time
goes into these projects and how much
people are giving of themselves
and so understanding that as you’re
using an open source project
being part of that community means kind
of contributing back and understanding
that
it only gets pushed forward if we all
work together
those are two great things that you’ve
learned over the years working with open
source communities
and just to summarize the first was that
you really have to speak the user’s
language and
the second that you have to remember
that we’re all in this together and
if you don’t work then you don’t
progress just to touch again on the
first point
i completely agree with you because at
the end of the day developers are
searching for good solutions for their
problems
if you can’t frame the problem from the
user’s perspective
then how do you expect users to quickly
understand and use your technology in a
way that you intended
i’d like to shift gears again and focus
more on the geospatial industry that
you’re working in
so firstly what does the geospatial
technology industry
do better than other industries yeah um
one just concrete thing to start off
with would be tiling
so in the geography and gis community
it was known very very early that
tiled maps were an efficient way to move
data from
the server to the client so we’ve taken
a lot of the
when we think of going from geo to other
domains
my last example with xray spatial was
going from other domains to geo
when we think of geo to other domains we
look at
tiling the handling of geometry
so if we’re handling points lines and
polygons
the geo community is very good at being
able to store and
analyze that sort of vector data and
then also
with handling continuous data in raster
formats
the geo community has come out with many
different formats
many that overlap domains but i think of
some great ones as being
you know hdf which came out in the early
80s to
netcdf and the geotiff the geo package
the shapefile
all of these data formats from the geo
world um
get applied in um other domains so those
are those are a few things from
from geo but web mapping has come a long
way
and uh geo has pushed
the use of geometry and of rasters for
mapping
to the web it seems like historically
the geospatial industry has been
quite advanced and ahead of other
industries in a lot of different aspects
but when thinking about the future where
is the geospatial industry heading
where do you see it in 10 years from now
i guess we we focus a lot on
on seeing what are the tools to scale up
and scale out
so as we take the the analyses that we
know how to do but scale them to larger
problems
that’s certainly something and as we
think about um
scaling horizontally we look at tools
like dasc
and what’s going on with dasc so seeing
other companies pop up for instance like
coiled computing
supporting specifically the das case we
can
look at coiled and say how could coiled
support not only the dash case
but also show integrations with geo and
as we
we look at um scaling uh vertically
we see a lot of the work that’s going on
at anaconda within number and the the
work going on at kwan site and open
teams with with number and ask
and see um a future where we’re able to
target
more hardware so we’re able to run our
algorithms for spatial analysis on
different um gpus or different types of
cpus or cluster configurations
and so those are about taking the tools
and
that we already know and scaling them
out so some of the things that i’m
particularly interested now
is our um distributed path finding so
like
distributed near optimal pathfinding
distributed view shed which is like line
of sight analysis and how can we scale
this
so that we can run it more often on more
data it sounds like the future for the
geospatial industry
is rosy and just like the rest of the
open source space i feel like we have so
far to go and it’s already grown in the
exponential rate and i think it’s just
going to keep growing
we’re aboard that rocket ship now and
there’s there’s no getting off so
thank you brenton and thank you pablo
for your time it’s been an absolute
pleasure chatting with you
and i know we’ll definitely have you on
the podcast at some point in the future
to those of you tuned in today thank you
so much for watching and listening
if you’re watching this on youtube then
please leave a like and subscribe to see
more content like this
and if you’re on apple podcast then
please leave a review
view it really does help us out to grow
the podcast and to get the open source
business name out there thanks again and
until next time
thanks you’re