00:03
ken
They say the sun will come out, but I still can't see past the clouds.
Jose Navarro
Someone's on my mind, yeah, I think I lost it, but I, but I...
Joseのキャリアの始まり
ken
Hello and welcome to our another episode of London Tech Talk.
And I'm Ken, your host, and this episode is brought to you by me and Kaz, software engineer based in London.
And I'm your host today because Kaz is busy and cannot make it.
And I invited an awesome guest today, so let me introduce him first.
Hello Jose, welcome to our podcast.
Jose Navarro
Hi, thank you very much for having me.
ken
Yeah, thank you. So I'm very excited with having you today.
I wanted to invite you to a podcast for long, even since the day we're launching this podcast.
So let me introduce to our listener first how we know for each other,
and then I want you to introduce yourself in any format you like,
your work, your career, your family or whatever you are excited to share.
So a brief history with me and Jose.
I met Jose when I worked for the previous company called Cookpad.
He's my friend, my colleagues, and also was my manager for a short period.
And I really enjoyed working with him at the time.
So I had a lot of great colleagues there, and Jose is definitely one of them.
And even after I've left Cookpad, he's my go-to person in anything from chat with from Tech Talk
and community building to busy parenting and family life.
So yeah, Jose, thanks for joining us. So can you please introduce yourself?
Jose Navarro
Yeah, thank you very much for a lovely introduction.
Yeah, so my name is Jose. I'm an MLOps engineer at Clio.
And yeah, as Ken mentioned before, we worked together in a previous role.
But if you like, I can give you a quick introduction about how I started in my career
and my journey from Spain.
So I studied in Madrid telecommunication engineer.
So for whoever who is not aware of what telecommunication is.
Basically, I was studying a lot of physics around how signals are transmitted.
So it's a lot of like electronics and physics around wavelengths
and how they propagate and things like that.
But when I when I finished my undergraduate,
I felt like I was missing a huge skill to have in the industry, which is English.
I wasn't able to I was able to read it,
but I wasn't able to have a conversation with an English speaker person.
So I thought with my girlfriend at the time,
let's move to England for six months to improve our English a little bit.
And then we moved to Bristol because we had some friends living here.
So it will make the transition a bit easier.
And then I worked in hospitality jobs for a while.
The six months end up being a bit longer because then I decided to join the Bristol University
to study a master's in machine learning topic that was quite interesting for me.
And yeah, once I once I finished,
then is when I started my career as a software engineer
because I wasn't planning to to follow a path in software engineering
since I studied telecommunication.
But once I finished my machine learning engineering,
machine learning master's,
then I found a job as a back-end engineer in a product company.
And I worked three years there where I learned a lot about good practices in software engineering,
both in back-end and front-end.
So after three years, then I felt like I wanted to become a data scientist.
That was my goal, right?
So I found a company which was doing AI in the defense space.
And then I joined them with the hope of becoming a data scientist.
But they hired me because they were struggling to productionize machine learning systems
into production because it's quite challenging.
So me with a background in machine learning and strong software engineering skills,
they hired me with the hope that I could bridge between those two teams.
And that's why I did.
And I found that that was much more interesting to me than becoming a data scientist.
So I worked there.
And then after that, I joined Cookpad where I met you as a machine learning infrastructure engineer.
And with the time, I ended up leading the machine learning platform team and all the platform teams at Cookpad.
And then after Cookpad, I joined Clio where I'm currently an MLOps engineer working in the platform.
ken
That's amazing. That's a very fascinating story. Yeah, that's a very fascinating story.
英語学習の経験
ken
So I have a couple of questions around your first six months you came to the UK actually.
So I didn't know that you had some goal for practicing your English at the time.
And it was surprising.
But most of our listeners are also Japanese or Japanese-speaking people who are interested in
or who are already working in an English-speaking environment.
When you came to the UK and you were working in the hospitality sectors,
and how did you brush up your English?
Is it like making a lot of friends and going to cafe or pub?
Or what was your effort there?
Jose Navarro
A little bit of everything.
Because I wasn't ever a very good English student.
I wasn't very good at studying the language itself.
So my way for learning was trying to get around of many non-Spanish speakers as possible.
So that was either like through work.
I was lucky most of my colleagues were from Eastern European.
So we had to speak in English or people from Bristol.
So I was quite lucky with my jobs in hospitality.
But then outside that, I tried to find hobbies that forced me in situations
where there was no escape of speaking English.
And that was true.
I had a group of friends who I cycled with.
So we would do like day trips.
So we would go on bike trips.
So there was no escape from that.
And also I joined a volleyball team.
Volleyball used to be another of my hobbies.
Now with age, I had to retire.
But also joining the team were like training sessions.
And also like weekend trips to play competitions.
That was my way of like learning with time.
But it took a long time.
I realized that my level was a lot lower than I thought when I moved to the UK.
ken
Wow, this is when you studied volleyball.
I didn't know that. Wow.
And is there any timing that you felt like,
Okay, now I'm comfortable making friends in English.
How long did it take for you to be comfortable with walking in English?
Is there any clear timing?
Jose Navarro
It definitely took a long time for me.
I guess it changes.
It's a very personal case.
It depends on what's your level when you start and when you arrive.
In my case, it took a while.
Firstly, most of my friends were either European immigrants in the UK.
Because it's easier to communicate with them.
Because kind of like the level is similar.
So there is no chance of having a very deep, complex conversation.
And so their level is also a bit lower.
Jose Navarro
So you can understand.
And the vocabulary that you use is fairly simple.
So it's easier to communicate.
So firstly, my group of friends were mostly Italians or Polish or French people.
Whereas once I started working as a software engineer after three years.
It took me three years to land my first job as a software engineer in the UK.
Then is when I started working daily on software.
That probably speed up my capacity to communicate in English.
And when I started making real friends from England.
That's cool.
ken
And now you're managing the team then.
This is great.
Jose Navarro
That's awesome.
ken
And I want to quickly shift focus.
To your course that you took at the University of Bristol.
And you say you took the Master of Science, right?
And how long did you go there?
Was it one year or two years?
Jose Navarro
It was one year.
Full time.
I did full time working at the same time in the restaurant.
Which was quite challenging.
ken
Was it intensive?
If you describe your one year.
Was it intensive one year?
Or is it more like relaxing from the work and enjoy studying?
Jose Navarro
What was that?
I lost almost 20 kilos.
ken
20 kilos?
Jose Navarro
It was stressful.
I was working about 25 to 30 hours.
As well as going to classes and then studying.
So there was a lot of time where sleeping three to four hours.
It was the norm.
But I didn't feel like I was struggling at the time.
Because the course was so interesting that I was enjoying a lot.
And I made a lot of friends that I'm still maintaining.
In fact, the CEO of Clio is one of my best friends from the course.
ken
Oh, cool.
Jose Navarro
And then I had to work to be able to survive.
Because I was paying for everything myself.
And I was paying rent.
My girlfriend, now my wife, was also working.
So I was lucky enough that I could reduce my hours to 20, 25 to 30 hours.
Whereas she was working full time supporting us in Bristol.
ken
Nice, that's a lovely story.
And this is also when you met the Clio founder, did you say?
Jose Navarro
Yes, at ML, at the Machine Learning Masters.
We were a group of like five more close friends.
One of them, Barney, who is the CEO of Clio.
He found Clio just a couple of years after we finished our master.
And it's been very successful.
ken
I remember it was 2016?
Jose Navarro
2016, yeah.
He created Clio around 2016, yes.
ken
Oh, that's nice. Wow, that's great.
And then after you graduated from the University of Bristol, you landed on the job.
After graduating the job, it depends on macroeconomy at the time.
How easy or how difficult was it for you to find the job?
After coming to the UK and after graduating university.
If you have a master's degree on machine learning, was it kind of easy for you to find good jobs at the time?
Jose Navarro
I wouldn't say it was easy.
In part because I was still struggling to communicate well.
So I didn't do well in interviews.
キャリアの初期経験
Jose Navarro
But also I didn't have much more experience as a software engineer before.
Because in telecommunications, we did some programming courses.
But it's not like a computer science undergraduate.
That has a lot more understanding of the technologies and building software.
So the master was good.
I did well in the master.
But I didn't have enough experience to land a junior data science job.
And I didn't have enough experience to also convince any company to invest in me as a junior software engineer.
So it took me a while.
But at the end of the day, I managed to get the first role.
And from there, it's been a lot easier.
Because it's much easier to go through interviews when you have prior experience building software.
Because you can talk about that.
Whereas as a junior, it's more difficult for companies.
ken
Yeah, totally makes sense.
And after you landed on the first professional job as a software engineer,
What was your focus in your career?
Like getting experience on writing some production applications?
Or did you focus on ramping up specific technology?
What was your focus in the first three years when you landed on?
Jose Navarro
When I started in my first role, I joined as a production engineer.
I wasn't involved in building new features in the product.
What I was responsible for is going through a backlog of issues that the customers reported.
And then trying to reproduce those bugs.
And then fix those bugs or help the customers to go around the bug while we were building it.
So I did that for a few months.
And it was great because it gave me the opportunity to have a wider view of the whole product that we had at the time.
Because as a production engineer, I was fixing bugs everywhere.
So I had the chance in a few months to be able to be exposed to everything in the product.
Once they realized I was quite successful at that, then they moved me into a product team.
And then I was more involved in the day-to-day building new features into the product.
And the technology for this product, it was a Django app.
So I was mostly doing Python.
But it was a small team.
So sometimes we would have to work on the front end as well.
Which was JavaScript and Angular at the time.
So it gave me a good overview as well of the front end space.
ken
That's nice.
And did you love a small team that you have to wear a lot of hats from front end coding to infra and Django application?
And how did you manage through?
Because sometimes I find it difficult to focus on lumping up specific skill if I dive deep into such kind of startup company.
When I joined startup company first time, it was like 2016 or 2015?
2016, I think.
I remember that.
I was still junior engineer and I have to learn everything.
I have to learn Rails.
I have to learn Postgres.
I have to learn Terraform.
I have to learn everything.
How did you manage that?
Jose Navarro
It's a challenge.
But I think it's a great challenge to have as a junior engineer.
Because it gives you so much areas where you can focus in the future once you figure out what is your passion.
What is your passion.
For me, at the very beginning, I was very, very junior.
So I didn't mind.
I had colleagues that were great software engineers and they were teaching me great software practices around building code, testing strategies, being very thoughtful on code quality, on readability.
ken
Like refactoring or re-architecturing?
Jose Navarro
Yes, exactly.
So that gave me very good principles of what good software should be.
And then all of those learnings can be transferable for any other thing that I did after that.
So I didn't mind that much that I had to do a little bit of everything because it was very positive for me.
And I felt at the time that I was learning a lot.
I didn't mind that I couldn't maybe at least at the first couple of years.
I didn't mind that much.
ken
I didn't have enough time to get my understanding on how certain technologies work.
I get that.
And I'm going to ask how, when did you get promoted to like senior level software engineer?
Because I'm asking it because some of our listeners are working as a junior or middle engineers and they are trying to get promoted at some point.
And what was your case?
And after, you know, working as a junior engineer, what made you get promoted to like in general senior level engineer?
Jose Navarro
When was that?
It's a very good question.
And in my case, it's a weird one.
I'm not sure if I ever had like officially a senior on my role.
Because I started as a junior engineer, production engineer.
And then after my three years, then I moved to another company to work more in the data science space.
And then that was software engineer.
MLOpsエンジニアとしての経験
Jose Navarro
And I worked there for a couple of years and the salary increased a little bit, but the title didn't.
And then I moved into Cookpad, which my role changed to ML infrastructure engineer.
So I started in the more like infra SRE platform type of size of things.
And then my role evolved into leading the team, as you know, becoming an EM.
But my role really didn't reflect the change.
I never cared that much on what my role title says.
I get why people sometimes they really care about the title.
Because it helps getting a representative salary if you move to the next role.
Because if it doesn't say that you are already a senior, then it's more difficult.
But I never had that.
And honestly, personally, I don't mind.
I'm lucky enough that in the space that I work as an MLOps engineer now.
But in principle, what it is, is helping data scientists to be more efficient delivering ML systems.
I am lucky enough that the sector is very keen on hiring that type of role.
So I know that independently of what my role title says, I am able to find a role in a company that I'm interested in.
So I am lucky enough that I can select of what type of company do I work for.
So in my case, the title, it's less important.
ken
I totally get what you mean here.
And I totally love the way you framed the title.
And especially when you're working for a startup or small size company, the title is very vague.
And what you care is not the title or naming on the paper or on the CV, right?
What you get things done is more important.
And it makes you visibility anyway.
So I really love the way you frame it.
リーダーシップの考え方
ken
And I want to also ask about the first time you step into the leadership role.
And am I right that this is at Cookpad you kind of like officially took on the leadership role?
Jose Navarro
Yes, it was at Cookpad.
So I think I was working at Cookpad for a year or a bit longer than a year.
We were a team of two.
And my lead got promoted to lead platform teams.
And then I had the opportunity to hire two or three roles to help me out building the platform and supporting data science teams.
And then I became the lead.
So yeah, it was a very easy transition.
My previous boss, she was very good at delegating.
So we were taking the team decisions as a pair more than she was the lead.
And then I was just doing whatever it was in the backlog.
So it was a very smooth transition because I followed the same pattern.
And even though I had another two or three people working with me in the team,
we all built the platform and made decisions as a team.
Obviously, I always had the responsibility of the outcome and the communications with upper management.
But the decisions I always try to share with the team and help them make the right decisions together.
ken
That's cool.
When I moved to the UK office in Cookpad from Tokyo,
I see that you're managing a great team and I see that you're doing a great job.
But I want to ask about leadership a bit more because it could be a tricky question.
But do you see leadership as a skill that you can acquire?
And when you started having two or three people that report to you,
even though it was a smooth transition for you, did you find any challenging bit?
Like you have to learn new skills or even within the smooth transition,
what was the most challenging bit for you, if any?
Jose Navarro
Yeah, it definitely needs some skills.
And I believe they can be learned.
It's easier for some people or it comes naturally for some people with more natural communication skills.
But at the end of the day, I believe that a great leader is someone who cares about people.
People development and the outcome of the team.
So those can be learned.
And anyone who is passionate about the leadership,
from the point of view of helping others more than the title or the salary,
to be honest, nowadays, individual contributors can earn more money than leaders.
So you don't have to become a manager to be able to earn more money.
So people who jump into management should be passionate about helping others
or making the teams work better as a team than as a group of individuals.
And if it doesn't come up naturally, there are great resources out there.
There is one great book called The Manager's Path.
Let me get the author correct.
ken
I think it's from O'Reilly with blue cover.
Jose Navarro
It's from O'Reilly. The author is called Camille Fournier.
And it's one of the best books that I read ever.
And it teaches you a lot of these skills,
basically skills of care about the people that you manage
and how to give feedback in an impactful way.
So if those skills don't come up naturally,
there are ways and lots of resources to learn.
ken
Oh, that's great.
Actually, I remember during one of our one-on-one at Cookpad,
リーダーシップスタイルの探求
ken
you recommended a book to me the other day, if you remember that.
So I bought it soon after the one-on-one.
I have one more last question about leadership and management.
So you said you're focusing on your members' personal development
is one of the key factors.
And so I want to know more about it.
What is your style for understanding the people's motivations
and people's career goal?
And let's say you have three members
and everyone has a different goal in career.
One is interested in getting more salary and money
and one is interested in doing a great job at workplace
and the other is interested in making impact
and get visibility or validation from other.
And how do you build a trust relationship with each of them
with having different goals?
What is your style?
Jose Navarro
My style is to help them out,
build a path themselves for their own career development.
I think it's important for a person to have an individual thought
about what they want to become or what they want to develop in.
And then my job as a manager
is help them materialize in that path
that they come up with.
Not myself pushing them into
oh,you are great communicating or you're very funny,
you're very popular,you should become a manager.
No,it's like what you want to do,
think about it and help them out figuring out
what they want to do to become
or what type of things they want to learn
and then build a path on it.
One of the things that we do
is build smart goals.
For those who don't know,
smart goals are small,measurable,achievable...
Did you say smart?
ken
Yeah.
Jose Navarro
Yeah.
ken
Time,whatever,yeah.
Jose Navarro
Yeah,timeline.
ken
It can be measured by timeline.
Jose Navarro
It's also like doing a realistic time scale
and then we build smart goals about those goals.
And as a manager,I try to create opportunities
for them to materialize those.
ken
Oh,that's great.
And I said the previous question was the last question,
but before we move on to the Clare part,
the actual last question is
管理職としてのモチベーション
ken
about your motivation as a manager.
What is your biggest motivation for you
to wake up every morning and go to work every morning?
There are so many roles as a manager
and as a software engineer,right?
You have to build code,you have to manage people
and what is the most exciting part for you?
I would say the reason why you get up
and go to the workplace.
Jose Navarro
I take a lot of...
It's very self-rewarding for me
to see people develop.
It's a very individualistic thing for me to do
because it's real.
I really feel that it's something that I do for myself
more for them
because I take a lot of joy
from seeing someone else succeed,
helping someone else succeed.
For example,taking one of my favorite people to manage
is junior people
because senior people,they're very easy to manage.
It's much more clear what they want to do.
They know exactly what type of things they have to...
You don't have to help them a lot.
Whereas for junior people,
they're still a little bit more...
They don't know exactly what they want to do.
So you need to show them different paths
and say like,OK,so this is what it means.
If you do these type of things,
you can become or you can reach out to this level.
When someone manages to get a promotion
that is under me,
it's the best choice.
It's much...
Honestly,I get more joy from seeing people getting promotions
than myself,honestly.
I love it.
And the same for taking the team success
to the company
and make it more visible.
I love that part of my job.
Like making sure that
the hard work that everyone is putting
is actually visible in the company.
And it's one of the hardest bet for platform teams
because we are quite hidden.
We don't build the product itself.
So we can...
It's more difficult for us
to link our success to a user story.
So when that really happens
and we manage to make our
successful projects,
achievements visible
and recognizable by the company,
it's the best feeling ever.
ken
Wow.
I really love the way you frame it.
And I feel that
because if you remember
that we went to the local pub
in London the other day.
It was several months ago.
So you were not my manager anymore.
But when I share my small success story
like launching my own podcast
or how I'm doing at my company,
you seem to listen to me
very enjoyfully and carefully.
And you seem like
it's self-rewarding for you
to see your people getting successful
and getting promoted more than yourself.
This is great.
That's awesome.
Cool stuff.
So I also want to know more about
the current...
Clioの成長とサービス
ken
your job role in detail
and what you're doing at Clio.
And I also want to know more about Clio as well.
So as far as I understand,
Clio was, as I told you,
was founded in around 2016
by your friend from the University of Bristol.
And right now,
it's a lead C company
focusing on financial services.
But it's in the US.
So it doesn't offer right now in the UK.
So as I cannot install the applications
in my mobile,
I'm not very sure what kind of product you offer.
So can you start talking about
the business side of Clio
and then also talk about your job role
and what you do daily at Clio, please.
Yeah.
Jose Navarro
So yeah.
Clio started in 2016
in the UK.
However,
very quickly when they launched in the US,
they became very popular
in the US.
And the business was like more than 90%
in the US.
So they decided to
close the UK market
for a bit and then focus in the US,
which has been working
really well.
So a little bit
about what Clio is.
Clio is a financial
fintech application
that users can install
in their phones.
They connect their bank accounts
and then Clio
can offer
help
on the finance
to help users
to become more
financial health
healthy.
And the main
business
is coming from EWA
product
which is an
earn wage advance.
So when people
connect to Clio
and Clio can see the
transactions, they can see the
salaries coming in and
the expenses, everything.
Users can
request an advance on their salary,
a small advance,
which is loaned by Clio
into the user's bank account
at 0% interest.
So we are
a subscription-based
financial service
with
the goal of helping
our users to
get out of the
pay-check-to-pay-check
life.
So it's targeted to
users, Gen-C users.
And then
one of the
benefits of Clio as well
is that
as well as the EWA product,
they can
have a meaningful
conversation with Clio.
It's not like a
traditional bank where you see the transactions
and then that's it.
Clio is able to talk to you
about your financial health.
We have
budget
tools where
you can create your
goals and
then Clio will
advise you from time to time,
like, oh, you're spending too much money on Uber.
Maybe you should take the bus from time to time
or you spend this amount of money in the last week
in Amazon.
To try to help you reach your goals.
So that
it's working
fantastically well.
Very high percentage of the
users, I think
it's over 80%.
They said that Clio has
helped them to become
better at managing their finance.
That's amazing.
And the company
is, as you said,
CBC.
We actually go for
an investment round in a long time.
I think more than two years.
I definitely haven't been
on one and I've been working for Clio
in two years.
And we recently became
profitable.
Cash flow profitable.
So it's actually
one of the companies
that have
better
financial metrics
as a startup.
So we're growing 2x year on year.
We have almost
300 million
ARR
every year.
And we're currently positive.
Profitable.
That's awesome.
So to the point where
we are considering not to go to
Series D and just wait
a little bit more
and go for IPO
ken
soon.
That's great.
Exciting.
MLOpsチームの役割
ken
Can you tell me more about
what exactly you're building
as an MLOps team?
Jose Navarro
What is your role?
Of course.
I work as an MLOps engineer
in the platform team.
So what we do is
we build infrastructure
to run ML systems.
And we
adopt MLOps tools
or we build MLOps tools
for data science teams
to make
the projects more
efficient so that they can
build new models
faster, more reliable,
more reproducible
and they can deploy those
models into
production and those services into
production
much more efficient
So having to build
a lot of the
scaffolding themselves, for example.
And yeah, it's
working very
well.
We have a
very good ML
and data platform already.
And we have been recently also
jumping as
second-man type of
constructual type
of collaboration
with product teams
when they want to build
something very complex
that is out of the norm
or when they're struggling to make something
that they have more efficient
or they want to
adopt one of the tools that we have
but the use case is slightly
difficult.
So we've run a few of those projects
and they have shown
a lot of success
so that we're currently
hiring
two more
プラットフォームチームの成長
Jose Navarro
data engineers or MLOps engineers
with the hope that they can
stay with us in the platform team
but being able
to have more capacity to run
ken
more often these type of projects.
That's great. So you're growing the team
and you're trying to hire two more people
and that's great. And I see you
already work for like almost two
months now. What is the most
you know, like something that you
are proud of
from you
you have built in the last one and
nine months?
Is there anything
that you want to brag about?
Jose Navarro
I mean, everything
like not
personally because like we
are a team of three
so it's not that I worked
in isolation completely
but I think in the last two years
we have done fantastic in the platform team
so one of the first things we
started working on
was adding
pipeline orchestrator tool for
them to be able to
build reproducible training
pipelines
as well as
experiment tracking services
Probably one of the
most impactful
projects that we did
it was building
a feature
platform based on open
source tools
so lots of the
classifiers
that we have
they financial services
so we need a lot of features
around users to
understand
how much money
we can loan them
and
to understand
when they will be able to pay back
and things like that
so those models
the models themselves are very simple
they not
very complex machine learning models
but the features are complex
so the services were
incredibly heavy against
the databases
because they were calculating the features
at real time
so what we did
is we build a feature
platform on top of
open source services
like Feast
and
we build it in a way that
the data scientists can define the feature
in a very easy way
to be calculated
either in batch or in real time
in streaming
and those features get
stored in
a database that is very fast for inference
but also
they are storing an offline
feature store
which is based on S3
to be
able to query historical
the historical values
of those features for like
retraining models
and that has helped
Clio as a whole to
be more
resilient
because on
on Fridays for example
it was very common on Fridays
we have big auto scaling on Fridays
because lots of people request these advances
on Fridays
and it was very common
it was more common
that we would like to
say that
on Fridays
we had quite a few
incidents from time to time
because those services were very heavy
and those databases become
a bit slower to answer
therefore the backend would like
time out
and they would retry
those
calls
which would make our databases
even more slow
so it was a challenging system
to have
and after building this feature store
and migrating some of
these
into the feature store
we managed to reduce
the latency of the
systems by more than
two times
we
モデルの自動再トレーニング
Jose Navarro
reduced the pressure on the database
everything
is much more resilient
and now those teams are able
to retrain
the models automatically
because they have the features
they have the orchestrator
so they can now
they have a pipeline now that retrains
the model every month
and it's able to check against
the model in production
and evaluate the latest model versus
the model in production
and decide if we want to run an A-B test
with a new model every month
which before it was like
a three-month project to try to retrain
a model so I'm very proud of that
ken
Wow, that's great achievement, right?
That's awesome
So during the process of building that
is only your team involved in?
Do you have any like
SRE team or like
production engineering
platform engineering team
that support your like underlying
NWS infra or help you
you know, introduce some
like residency technique
Jose Navarro
We are the
platform team, so we are
the ones, yes
it's, we are a team of
the platform team is a team of
seven, but we are focused on
different aspects, so
data and ML
it's three people
It's two colleagues of me
Paolo and Fabio, and then myself
Paolo being
the staff engineer, so he's the most
like technically
advanced
and then Paolo
is coming from
a data engineering background, but
he's an amazing software engineer
and he's like
onboarding all the ML
systems super fast
and then myself
and we own both the
ML and data platforms
so like everything
from the Kubernetes
cluster and the networking
Clioでの経験
Jose Navarro
in AWS
to the
Redshift data
warehouse data lake
and
query layer on top of
S3 for example
so it's all managed and
operated by ourselves as well as
developing new products like the feature store
ken
so it's a busy team
Oh yeah, busy team
and I know you
change your industry from global
recipe sharing platform to the finance
you know, that Cookpad was
having the Rails application
for showing the global
and do you see
any gap in that perspective
like from the business, you're changing
industry to the
financial tech and how
much gap do you see
as an individual ML
Jose Navarro
engineer
ken
Probably not that much
Jose Navarro
Not that much
We
obviously working with financial
data, it's different from
the piece
but to be honest
that is all up to
what type of data can we
use to train models
Like for example, it's illegal for us
to train models based on
gender
So we know that
So it's not like
it changes from the
infrastructure point of view
Obviously we have
policies around
like we can
have like user data
on laptops or like what type of
transactions we can see and things like that
But from an operation
point of view, it hasn't changed that
much and Clio is
architecture is fairly
similar to Cookpad in the way
that the backend is also a Rails monolith
So
from my point of view
nomuch
to be honest
ken
That's cool, Sal
Yeah, and that's awesome
And actually we were thinking
about the third topic which is
I wanted to ask about the Bristol ML Meetup
but we already talked about
like 15 minutes and it was great talk
So I was thinking to
if you're fine with that
record another podcast someday
and so that we can focus on
Bristol ML Meetup and community building
And so for today
before
in the last 5 minutes or so
we gradually
get into the closing
and
is there anything that you want to
you know like share to our
コミュニティとネットワークの重要性
ken
listener about Clio
or your personally or you can also
you know share about your next
coming ML Meetup, is there anything
that you want to advertise, share
or promote, anything you want
Jose Navarro
Yeah, totally
So yeah
So from the Clio perspective
I
feel like Clio is a great company to work for
It's really fun
We're working on something that is impactful
for people
and it's
the first time that I'm working for a startup
that is actually
making money
So that's a very good
ken
feeling
Jose Navarro
Yes
We are expanding the team
So anyone who is interested
and is
around
in the UK really
because we give the possibility of
working remotely for Clio
Reach out
or have a look on the website
Clio Jobs, if you Google
Clio Jobs
C-L-E-O
Then
There is tons of jobs
opportunities at the moment
and
what else I wanted to say
Yeah
ken
How do
How do people
can find you online
You can
drop the link to any SNS
if you want to on our website
and on a show note
if some of our listeners are interested
Is there any way to reach out to you
Jose Navarro
Yeah, I'm happy to share
my LinkedIn
profile for you to post
on the description of this podcast
I'm always happy
to have a chat about
ML Ops, ML
career development as well
if someone has any question or wants to become
this or that and they want to
have a chat
Some people have reached me out in the past
I'm super happy to set up a call
and have a chat about
any topic
So yeah, I'll give you my
details and I'm happy for you to share
ken
That's awesome. Yeah, thanks for sharing
and nothing more
That's cool. Yeah
All right. So yeah, let's get
closing. So I really appreciate
you being with us on London Tech Talk today
So thank you very much for coming today
and sharing a valuable insight
your career history and
Clio, what you're doing and I really
enjoy talking with you. So let's record
another podcast which is going to be
focusing on your community building
aspect and ML Ops
meetup. So I hope
to see you again on another episode
and also for our listeners, thank you
everyone for listening and we will see you
next time. Take care
Jose Navarro
Thank you very much for having me. It's been really fun
ken
Thank you
