How We Hatched: Mingsheng Hong, Co-founder and CEO of Bluesky | Ep06

In this episode, our host Tim Winkler speaks with Dr. Mingsheng Hong, co-founder and CEO of Bluesky, a startup on a mission to make data clouds cost efficient. Prior to this role, Mingsheng was head of engineering for Google’s ML runtime (e.g. TensorFlow). Mingsheng also had a decade of experience building and delivering Big Data products and solutions at Google, Hadapt, and Vertica.

During this episode, Mingsheng takes us along for an inspiring account of his professional journey, sharing the lessons he learned along the way.

He shares:

• How his time at Google and Microsoft equipped him to found Bluesky
• The pros and cons of working at a large corporation vs a startup
• The process of raising a first round of funding as a startup
• Open roles on Bluesky’s team and the qualities they’re looking for in a new hire

Welcome to the Pair program from hatchpad, the podcast that gives you a

front row seat to candid conversations with tech leaders from the startup world.

I'm your host, Tim Winkler, the creator of Hatchpad, and I'm

your other host, Mike Gruin.

Join us each episode as we bring together two guests to dissect topics

at the intersection of technology, startups, and career growth.

All right?

Yeah, I'm looking forward to getting this story out.

Um, so we'll just kind of kick it right into it, uh, and, uh, uh, jump right in.

So, Ming Shang, thank you for joining us on the PAIR program.

It's great to have you.

Thank you, Tim.

Excited to be here.

Excellent.

Yeah.

And for our listeners, this is another bonus episode of a mini-series

that we call How We Hatched.

Uh, so this is a fun discussion to hear a little bit more about your unique

career journey, you know, where you kind of came from, how you arrived at this

current point, and your seat today as the c e o and co-founder of Blue Sky.

Um, so I always like to start by having you provide the

listeners with the quick overview.

Of blue sky and the the problems that you are solving here.

Sounds good.

Thank you, Tim.

So at Blue Sky we focus on making data clouds cost efficient.

So to unpack that a bit, data clouds or the cloud-based platforms that

provide data processing support and increasingly also ML support.

And one of the pain points for the last five to 10 years of

the world transitioned into.

Is that people need to make sure their cost efficiency by moving the workload

to data cloud can be as high as, or ideally even higher than on-prem.

And unfortunately, unlike what some of the cloud vendors have been, uh,

marketing, it is not always the case.

And that's a combination with the limitations of the cloud platform as

well as the best practices of how people use the clouds, uh, yet to be shape.

This is where companies like Blue Sky coming and we're looking to help

the world finish the transition to the cloud by providing the tooling,

the expertise, the recommendations to make the workloads cost efficient.

So that's where our mission at Blue Sky.

We started earlier this year and, you know, happy to talk

about our journey so far.

And maybe also the, you know, my personal journey prior to,

uh, leading to this point.

Yeah, very fascinating.

I mean, obviously a space that is just, you know, getting bigger

and bigger, uh, each year, uh, more and more data coming out.

And, you know, one of the taglines that jumped out to me on your, on

your all's website was, you know, your your Snowflake co-pilot.

Um, and so obviously with the rise of these cloud data warehouses like nof, You

know, this is gonna be something that, you know, uh, companies are gonna want to

adapt to, uh, just to optimize and become more efficient, uh, when they started

gathering more and more of this data.

Um, so let's, let's, um, uh, you know, let's go back in time a

little bit here and tell me a little bit more about, you know, where

whereabouts you, you grew up and how you got into the, the, the world of.

Sure.

So maybe let me just go it backwards chronologically.

So, before Blue Sky, I spent close to nine years at Google as an engineering lead.

I was fortunate to have led multiple teams, uh, building out some of the

mission critical software components, such as the storage and querying layer

underneath the hundred billion dollar.

that's where it was very exciting, but sometimes nerve-wracking

to see every transaction going through my code and my teams code.

And fortunately we were able to make the operations pretty smooth and we managed to

make the backend much faster and cheaper and turn improving users and experiences.

That's who doesn't like, you know, having more interactive, uh, dashboards that

are supported by more performance back.

. And after that I got a chance to move to Google's AI division, Google

Brain Team, and was fortunate to lead, lead, lead the team to take,

uh, the first version of cancer flow.

As many of you might know, it powers Google's internal

mission, critical AI workloads.

And it was built by Jeff Ding and others.

Jeff was known as the engineering guard at Google, and we were fortunate

to step on the shoulder of the giants like Jeff, learn about the work.

The new challenges and we took the backend, the potential flow

run time to the next level.

Revamped it, launched it, uh, later, last year and was able to gain really

big impact by making mission critical AI workloads like ads faster and

cheaper, and in turn allowed them to reclaim the resources that data center

resources to then invest into new business features, launch new features

to improve ads quality and improve Monet.

So that's one key insight I also gained from the cost efficiency

journey where we are not only improving just the bottom line, but

by helping people reclaim and use the resources in other ways to monetize.

It also improves the pipeline, and I can go back further.

Uh, before, uh, the Google journey, I spent five years working on, uh, early

stage enterprise database companies.

That was back in.

So there I learned a, a ton by being an engineer in early stage companies.

I got exposed to not only working on engineering problem, but also being

customers, uh, facing, uh, supporting and up, uh, upselling, uh, deals as a sales

engineer and also doing product marketing work, tech evangelism, like giving

talks at large scale conferences like

And that helped me build a foundation to appreciate.

Critical elements needed to make enterprise software successful.

Beyond having good tech and engineering products, all of the sales, marketing

and other elements need to come together.

And before that, I spent five years pursuing my master in

PhD degree in computer science specializing in databases.

So when people say, oh, this database problem is really hard, we need a PhD

to figure it out, I'm happy that at least I got that dimension covered.

Before that I was born and grew up in Shanghai and the first 24,

uh, 22 years there finishing my undergrad degree in computer science.

So I'm a first generation immigrant, decided to be here, uh, pursuing my

American dream, uh, as an entrepreneur.

Really living the dream, everything.

That's wonderful story.

So you, you came over when you said you're, uh, in your twenties, um, and,

uh, where did you come into the States?

Were you out West coast?

You, you said you, you touched on Boston.

Uh, whereabouts were you located when you first came into the States?

So, uh, about when I, uh, was about to finish my undergrad in Shanghai,

I applied to various US grad schools.

I received a few offer.

And was fortunate to get a very good, uh, scholarship from Cornell.

And I like Cornell, not just because it's one of the Ivy League schools, very,

uh, reputable, but I believe it also has the strongest computer science program.

And I especially liked the database group.

My advisor was Professor Johan Gil, the, a very strong, uh, uh, you know, scientist

and also a well-run manager and a great.

Currently Johans is a research director at Microsoft Research Redman Lab.

I'm still, uh, in good touch with Johans and he's actually one of our angel

investors and advisors to Blue Sky.

So really fortunate to have met with great people, mentors like Johans.

And kind of build out my computer science and database foundation.

I love

to hear how these folks come, come back into your life,

uh, you know, later, later on.

You know, these is one of the first folks that you interacted with

when you came into the States.

And, uh, you know, now, uh, he is participating as an investor in

your, in your current ventures.

That's fantastic.

Um, so, and you, so, and looking at your profile, you spent a little

bit of time at Microsoft as well.

Yes, I visited.

Microsoft Research, I believe three summers in a row.

When I was a grad student, I always loved each visit.

So when I got invited back, uh, you know, out of just, you know, continued

that having more good time at work.

And I loved the summer at Seattle, so I didn't really, you know,

uh, I didn't really then go and, uh, look at other places.

There are other wonderful places as well, but I love the three summers, uh, at m.

Learned a lot, uh, worked on cutting edge technologies like extending Microsoft

SQL Server with stream processing.

And that happened to also be my PhD thesis topic.

And it's great to see how innovation can get landed in industry setup.

And that's also one of my motivations for them getting into a startup as the first

segment of my journey after graduation.

Yeah, that's fascinating.

Um, you know, the first startup that you, you joined with, um,

first of all, are they still around

or.

Yes, the company was founded by a professor named, uh, stone

Breaker, Mike Stone Breaker, with a co-founder c e o, uh, Andy Palmer.

And by the way, Andy is also one of our devices at Blue Sky.

It's really fortunate to continue to connect with them and, uh,

get their supporting insights.

Professor Stone Breaker has about four decades.

Foundational experiences and contributions to building out SQL databases.

And because of that, he received a touring award known as the

Nobel Prize in Computer sciences.

And I was fortunate to be able to work at one of his companies.

Uh, he moved from Berkeley to m i t in the early two thousands, and he founded that

company out of the MIT Research Group.

Mm-hmm.

. And so that was my first segment.

That company was later acquired by.

In 2011, I joined in 2008 with the anticipation, uh, out of a naive

young graduate student that it'll become the next hour to speak

because I read its research paper.

I still distinctly remember back in 2005 and I, I read follow up work, uh, in 2007.

I am a big believer as technology can, uh, can leaf the incumbents

by being 10 x faster and more.

, but little did I know in order for a company to be successful at the

scale of replacing the incumbent, there's a lot more to just having the

cutting edge technology and product.

So that was my four years learning at Verica, where after the first

three years, it was acquired by hp.

Everybody was super excited about me, but frankly, I felt quite disappointed just

because of my own expectation being set.

But the silver lining is then after the acquisition.

I got a wonderful one year, uh, tour of being customer facing, uh, and going

as I mentioned earlier, working with the big radical customers like CIGA and

others where I deeply understood, uh, the workloads, was able to demonstrate our new

product features, help solve new business problems, upsell, uh, closed deals.

So that also generated a new sense of.

part of my work, and that's how I learned to talk in customer's language and

present complex technical ideas in a way that can get understood and appreciated.

And that's the foundation I take to my day.

Uh, day-to-day work at Blue Sky.

In a

field.

Yeah, it's a fascinating profile because you know, you don't often see, you know,

so many engineers that that kind of border this line of product marketing.

And I think that's something that, you know, through this experience and

through this story, you know, you can start to see how product marketing

and how vital that becomes as a part of, of a business and their growth is.

And really being able to relate to those customers and to, to get

that opportunity in such a large organization like that, I think is, uh,

uh, it's a fascinating, uh, you know, piece of the story that stands out.

Um, so you, so you moved to, to Google and, um, you know, I'm just

curious, you know, how did you come across the Google opportunity?

Was it somebody from your past life that, you know, was over there and said,

you know, Che, you gotta get over here.

You know, we're, we're, we're doing some really interesting things.

How did you, uh, settle on Google as that next place that you wanted

to, uh, you know, spot your career?

Indeed.

It's funny that, uh, so tracing back to our grad student time, uh, I went

to Microsoft Seattle for internship.

Some of my friends went to, uh, bay Area, uh, Google for internship

around 2005, 2006 time, and some of them never came back.

That didn't really, uh, I didn't really dig more, uh, much into it.

I knew there were good places in the valley, but I'm very focused

on finishing my PhD study.

And in 2008, I got a very good offer from Google and for family reason,

I had to stay in the East Coast.

So I was thinking about joining Google, New York.

But then I went into this wonderful vertical opportunity where out of the

company's eight engineers, everyone interviewed me and even sales and product

management inserted themself into the.

So the interview was super exhausting, full day, but I was very impressed and

that was one of the reasons I chose, uh, radical over Google at that time.

Naively thinking that I can always go back to Google any day.

And that , that prediction came true.

And so, uh, back in 2013, I've been in touch with my friends who joined Google.

Earlier on, they told me about all of the wonderful innovations, the

engineering culture and everything there.

And so that was a place that I really had positive impress.

And so back in 2013, an opportunity opened up, uh, uh, an old friend and,

uh, you know, Cornell classmate of my office mate of mine, uh, helped me

connect with their vp, uh, of ads who runs the ads and data based backend.

So I was recruited, uh, at a pretty senior position, uh, and I was able to leverage

my, uh, past academic background as well as industry background at, uh, radical.

And adapt to then make the red contribution to Google's ads back backend.

As we mentioned earlier, uh, I led the team to basically revamp, uh,

the background, the back backend for storage and querying, upgrading the

storage system from row based to column.

And back then I thought column there upgrade was a piece of cake, so to speak.

As I have done this, I have seen it how it's done at Red account.

It turns out at Google's workload in terms of.

And requirements, it's very different.

So there's a lot of new learning for me.

And uh, in the end we were able to finish it, not just completing the

product, but also migrating all the existing data in about a thousand steps.

So the analogy there, as people uses flying the plane while changing the

engine, one little piece at a time, and fortunately, uh, the migration was

very smooth and this was pretty scary, even in hindsight because changing the

storage format means data might get.

If data is corrupted, then ads users will be very unhappy.

And so I'm very thankful for the leadership support and trusting

our team in this piece of work.

And also through the hard work and a bit of a luck, we were

able, able to pull it off.

So very proud of

that achievement.

It's gotta build a bit of confidence for you, uh, as an engineer, you know,

coming in and, and getting that level of.

You know, from your team to, you know, go, go forward with this

data and, and ensuring that it's, you know, all quality driven.

Um, your, your title when you come into Google is this tech lead, um,

supporting analytical data infrastructure.

Is that, is that right?

Yes.

And you're in this seat for, looks like four years, um,

and then you end up mm-hmm.

uh, transitioning into, um, the tensor flow, uh, kind of environment is that,

And then one of the big things that jumps out here, You know, in 2019, um,

you know, your, your role it looks like becomes this head of engineering, um,

within the, the TensorFlow department.

So, uh, we talk a lot on, you know, on this podcast about,

you know, FANG environments.

Um, obviously comparing some of these to startup environments.

Um, but I just wanna touch real quick for our listeners that are maybe navigating

career growth within FANG environments, you know, what are some of those things

that you would recommend to, to folks?

You know, maybe, you know, starting their career off in a large, big

tech organization, um, and wanting to navigate that career journey.

Anything, anything that you recall, uh, that would be beneficial to folks

that are, you know, looking to, you know, build their career growth,

uh, within those FANG environments?

Yeah, sure.

So first of all, a quick clarification.

I was running the TensorFlow backend called Runtime.

So I was head of engineering of TensorFlow Runtime.

TensorFlow itself is a larger organization, uh, where my manager

was one of the senior leads for TensorFlows engineering.

Uh, so yes, I did have a fair share of my experiences having spent close

to nine years at Google and earlier a few beautiful summers at Microsoft.

So I can talk.

, um, all of the great things I appreciate at a place like Google and how that

contrasts with startup environments.

Mm-hmm.

, uh, first of all, Google has a lot more resources and, um, that

translates to, uh, 10, you know, the tendency to give engineers more

room to innovate, to experiment, and, uh, sometimes make mistakes.

The management culture can be more encouraging and that is great.

It's also more hands.

. So for certain type of people, uh, it might be senior people who would prefer

more autonomy, but they could not also be junior engineers who would prefer to

just, you know, try out their things more.

With little guidance.

This could be a pretty good setup.

So in my experience for, uh, Google, there are two kinds of engineering projects.

One is pretty incremental, uh, you know, for such a.

Operational base at Google, making any change is non trivial.

It requires a lot of, uh, efforts in understanding the existing design

code base and thinking about how to carefully extend the existing code

base with the new feature, how to test it and launch it carefully.

Now, sometimes, as I mentioned, involving dozens or even hundreds of

steps, so even a minor change might take quarter, if not years of effort.

These efforts are very rewarding in terms of the end impact.

Even though it may feel tiny because of the large operational base at

Google, let's say hundred billion ads, if you, someone is able to

improve the efficiency by even 0.1%, that is 1 billion worth of impact.

So impact wise, it is great.

It depends on the engineering taste.

Some engineers would prefer to tackle larger scope challenges.

In that case, this kind of project may not be a good fit.

And then the other kind of project is more, I guess I could

call it blue sky, like project.

No, think bigger experiment with things.

Don't worry about the operational stack.

Try to bull shop something from scratch.

And as I mentioned earlier, since Google has the resource to sponsor

this kind of risk, Projects, uh, some engineers almost working in the research

capacity enjoy this type of freedom.

Of course, now with the down economy, it seems various fan

companies are tightening up.

You know, the leadership ring a little bit.

I understand that.

But in the good times, uh, such projects, were pretty, uh, well supported.

Many of these projects fail in.

and Google still supports them.

The engineers can continue to, if not getting promoted, at least continue to

get the same great salary benefit and continue to experiment and innovate.

And then occasionally you have a project like the, uh, ads backend, the column

store I mentioned, uh, it was productized into a backend system called Mesa, m

e s a, and it's then carried over to the next generation system called Napa.

It's like Napa Valley.

And there were research papers in.

on these subjects as well.

So that's the best of both words.

We get to innovate from ground up and in the end to hard work and a better

of luck, uh, impact the production, integrate that into the production.

So these are very satisfying experiences.

Mm-hmm.

. Now one thing I would say, just to balance the arguments a bit, is in

contrast to startup, uh, sometimes the workplace may be a bit slower.

I already mentioned one reason, just the complexity of

understanding the existing stack.

Trying to understand it, uh, extend it properly.

The other factor is just the human complexity, the organization, company.

There might be more stakeholders we need to align with, and that's super important.

And if you know, I would be happy to share, uh, my later

journey in the cancer flow.

Uh, organization especially that kinda highlighted some of those challenges.

Yeah, the human complexity is the, the, the area that we

like to kind of pick apart.

Um, we, we, we say a lot of things that we work on here at Hatch and, and at

Hatchpad, the community focuses on the, you know, this human side of engineering.

Um, you know, when you would, when you move into a startup environment,

um, you know, things are gonna be very, you know, fast moving.

You might not have all of those resources that maybe you could have leaned.

At an environment like Google when you, when you made the transition to Blue Sky,

let's, let's jump into this now cause I think this is a good, uh, pivot point.

Um, what would you say was some of those instrumental and valuable things

from Google that you would say, like on the human side of things that

played a part in you as a founder, um, and, you know, getting the trust

of folks that are joining your, your team in those early stages.

, what was it that you say, you know, was, was instrumental or, or helpful for you?

Yeah, so a couple elements.

First of all, it is the credibility we carry.

Uh, my co-founder, John from Uber and myself from Google.

So given the track record, uh, we earned, uh, what we were able to earn respect from

our funding team members through, uh, the interview process and then the day-to-day.

Another part is the engineering culture.

The practice of, for example, thinking from first principles.

So we try to learn from others lessons, but also when it may not,

uh, be aligned with our intuition.

We are not afraid of digging deeper and trying to derive, uh, so for

ex uh, trying to derive the truth.

So for example, uh, conventional wisdom can.

Well by moving workloads to the cloud, since the cloud vendors have

a lot more expertise and they have economy at scale, of course the cost

efficiency should be higher in running your own workloads, OnPrem, right?

But the reality isn't that simple.

With cloud, there are new challenges with how to use them efficiently.

We have seen, for example, from one user workload, a single

employee could accidentally consume $10,000 worth of credit.

Without generating any business value.

This is because all their queries actually failed and

maybe they were not aware of it.

So you see these credits being burnt with no value and the head of engineering

head of data did not have visibility.

So that's one of the examples where, uh, they were not sufficient best

practices or God wills installed to basically provide the proper governance.

I'm sure at the same.

if an employee ha, you know, had, uh, goes and have a business launch that exceeds

a threshold, let's say a hundred dollars, they need manager approval, right?

But so how can they spend $10,000 of compute credits with no questions asked?

So these are the exciting, uh, these are the problems that excite us.

Uh, we, we want to figure out both on the technical side, but

also on the organization, uh, the human side, the process side,

how to, uh, improve the process.

So from first principle thinking, then there's also being

curious and learning driven.

Uh, and while we have learning share with others, and, uh, there might

be lessons we learned from negative experiences and we want to carry over

good culture practices from Google.

Like, uh, blame this postmortem.

So these are the, you know, the, the good elements we carry over to Blue Sky.

Hey, startup Techies has this podcast inspired you to explore

a new startup career opportunity.

Then make sure to check out my hatchpad.com/jobs to browse

startups by stage, tech stack

and salary.

So with Blue Sky, what is the, um, the current headcount?

We have about, uh, 12 full-time members.

Mostly in the r and d department, product and engineering, but

we also have exceptionally strong and yet small business.

And, um, you all are, uh, you have some seed funding, is that correct?

Yes.

And you're backed by Graylock?

Yeah.

Which is

a LED by Graylock, uh, with, uh, support fund additional

VCs and, uh, enjoy investors.

and Greylock, you know, obviously a, a very well renowned venture firm.

Um, you know, how many meetings did it take you before

getting Greylock to invest?

I'm, I'm always think curious from a founder's perspective, you know, this, uh,

this journey of GA of getting investment.

Um, talk me through that and why do you think Greylock, you know,

decided that Blue Sky is a good.

Great question.

So let me try and speak from, uh, a Greylock perspective, but with

a caveat that maybe this is a question better for our, uh, board

member, our investor, Jerry Chan.

Jerry is wonderful.

We had a small number of meetings before having this handshake

agreement, and we were fortunate.

I think one of the reasons for the low overhead is we have had prior

relationship with the Greylock investors.

So they kind of, uh, knew, uh, uh, the past work for a number of years.

And, uh, for myself, I also got a strong, uh, intro to another

Greylock investor back then.

Sarah Go, Sarah, uh, started his, her own fund recently, but we are

still getting great support from Sarah, of course, as well as Jerry.

And, uh, basically through the, through the intro and through

the past work, basically the, uh, engagement started at a.

High level of the foundation.

And, you know, you, you mentioned about 12 folks.

Um, most of these folks, uh, you know, are, are engineering folks, tech folks.

Um, I noticed on, you know, on uh, LinkedIn, a lot of these folks come with

the title of like founding engineer.

Um, this is something that, you know, we've done a lot of research on.

It's one of the hardest.

Positions to, to recruit for.

Um, you know, what is it that you would say was helpful for you all

with recruiting this first round of engineers in those early stages, um,

and the, the title of founding engineer.

And I know that sometimes titles can, can mean something different to everyone.

Um, but what would you say differentiates a founding engineer

from maybe just like a, a senior

engineer?

Ah, that's a great question, Tim.

we subscribe to the culture where the company initially

shouldn't have super fluid titles.

Mm-hmm.

, when we give people unnecessary titles, it might generate more problems.

So we try to minimize that.

So we don't really have levels or levels that indicate the

engineering, uh, seniority.

This might be different from some other startups, but that's.

Culture we subscribe to.

So at Blue Sky we only have titles for customer facing roles where it's useful

for customers to understand, uh, the role of the individuals they work with.

But then internally we would want to just everyone work in a pretty flat

way, and it's meritocracy, merit based, uh, when we debate ideas.

And so for that reason, we don't really have levels like senior versus support.

Instead, we use words that are more factual.

If they joined when we are still in the funding stage,

then they are founding engineer.

It's just a fact.

It doesn't otherwise differentiate them from the later engineers who join.

And someone can be an area lead, let's say marketing lead or a

deployment engineering lead.

And the uh, the lead is also a fact.

They're currently leading that department.

But otherwise we.

At this point provide different levels of director versus vp, etc.

Mm-hmm.

, that type of titles.

That's fascinating.

Yeah.

It's, it's gonna be the type of thing too, as the company evolves.

You know, look, thinking through, um, you know, the, how the org uh,

specifically, you know, when we, when we talk, we like to dial in on like the

engineering teams or the product teams.

Um, you know, have you thought through, you know, uh, when is that le level

that you're, you're needing, uh, ahead of engineering or, uh, you know, a

director of engineering and, and you know, what is it that you know comes

into play for you that, that sparks like, this is probably the good time for.

So our view is initially, uh, the engineering team needs to be,

needs to have a, a number of, uh, very strong individual contributors, tenex

engineers and generalists, uh, and have a kind of culture of being scrappy in the

sense, no, let's say when I walked at.

Before we start any large scale effort people or basically condition cultured to

write large design docs, because that's also part of, uh, a big way of how they

get evaluated in their performance, but that's not how startups operate.

In the end, we prioritize on getting things work, and then along the journey

of course, we need to emit, you know, just enough communication bits, be it in

person or write a short email or slack or.

To make sure we get enough alignment, support from other stakeholders,

but then we don't, for the sake of writing a large design doc, we write it

So, uh, it's important to have aligned culture where we can move fast.

And making correct decision initially is important, but it's even, even

more important what we get, new data we can very quickly course correct.

And so these are the, the traits, soft skills, the culture.

Uh, we seek ground funding engineers when the team skills.

At some point we might need a engineering manager, but this, this point we'll

prefer to have a flat hierarchy cuz there isn't too much management challenge.

People are not asking for performance review, even though we deliver

that in a continuous realtime basis when we give feedback.

But otherwise we don't need the formality of.

A performance review cycle, which in my experience could be

consuming a lot of time and efforts.

Yeah.

So these are couple of things, but at some point as we still, I'm sure

we'll bring in engineering manager, director of engineering and so on.

Yeah.

I think it's a great perspective.

You know, when, when we, when we talk to founders in these

smaller stages, you know, you don't want to get too distracted.

Um, you know, the, the org chart at large, you know, how is it going to evolve?

But more so let's build, you know, let's focus on our customers.

Let's be scrappy.

I think this is a term that gets, gets tossed quite a bit and I, um, I

can super appreciate that, you know, when, you know, building a small team

over here at Hatch as well in those first, you know, five to 10 folks,

you know, we were heads down, we were.

Talking about how to generate revenue, we were really focused on our customers.

And, um, you know, there will become that time to, to think bigger, you

know, org evolves and, uh, you know, layers and things of that nature.

But right now it sounds like you all are very heads down and, and

that's, uh, where you need to be.

Um, what is the, um, you know, what is kinda like the core tech stack

of, of what you all build with?

So we build on the public, Mostly aws.

We also use, uh, G C P and, uh, we will of course want to expand to

Azure, uh, and maybe other public clouds in the future as needed.

And, uh, basically we have e TL tools like, uh, air Byte to extract,

uh, metadata from customers.

Snowflake, uh, instance, and this is the metadata, like query history

that we use to do our own a.

And make optimization recommendations.

So that's our sequence source or core it.

So there we write our internal logic, uh, in Python, uh, SQL and

so on to transform, enrich the data and uh, make recommendations.

And then we have a native UI framework, uh, to then surface the visibility, the

recommendations, uh, to the end users.

. And

is the team fully distributed or whereabouts are the team?

Uh, the tech teams kind of located,

we're about half, half, so half of the team is in the Bay

Area, but uh, we do recruit in a pretty wide range right now.

Folk folks are basically foc uh, concentrated in the US

time zones, including Canada.

We used to have an engineer in Asia, uh, a contractor, but then it's been a pretty.

Experience logistically to coordinate the meeting, especially then with New

York, uh, and with, uh, bay Area.

So for now, we're gonna focus on just the US time, time zone, and, uh, we'll

continue to be opened, uh, for remote hire, although we will continue to,

uh, make sure there's enough in-person time, like back in September we had a

nice, uh, all hands, uh, team gathering where we flew everybody to the bay.

And would want to do that on a regular basis, not just in the Bay

Area, but find some other good, interesting spots to fly people

together.

Sounds like you really value that, that culture.

You're bringing folks together, making sure folks know who, who

one another are on the team versus just, you know, another day at work.

Uh, this is something that you

value.

Yes.

Uh, especially in early stage startup, uh, business.

, it helps for people to build even stronger trust so we can iterate faster

and we need to, you know, pass along new thoughts and, uh, tweak on the directions.

So it's important to have enough of the, the bonding and I think it also

improves the personal satisfaction.

Mm-hmm.

that engagement as we work together on a shared mission, uh, being

able to more deeply understand, know each other, uh, improve sfa.

I get, um, I wanna ask a little bit about the, the openings that

you're actively hiring for, and then we're gonna jump into a, a segment

called the Five Second Scramble.

But before we do that, what, what are some of those key roles that

you would say that are high priority for Blue Sky, uh, on the tech front?

So at this point, we are looking for a senior product lead.

Ideally someone who has downstairs work before taking a product from zero to.

So understanding how to do user interviews and then select design partners and

integrate with them to continue to enhance our product, adding new product features.

Uh, the founders, we had pretty good, uh, pro uh, technology background.

We also worked with product managers in the past companies, so we are kind

of covering the product management for the time being and with the

support from the very junior, uh, product manager currently as well.

But we're looking for a senior product.

And in addition, we're looking for, uh, another 10 x engineer.

Uh, they could become our chief architect or engineer, just a very important tech

leadership position with a strong being, a strong, uh, individual contributor.

And also showing by example and leading the team on the implementation front.

And in addition, a couple senior engineers across different, uh, infrastructure,

uh, function functions like DevOps.

Data pipeline infra to help us prepare for scaling up to supporting dozens and

hundreds of customers down the road.

Sounds like an exciting time to join the company.

Um, the current team that's in place, uh, their backgrounds, are they a mixed bag

of big tech and startups or, you know, do they tend to have more of a theme of a, a

background?

It's pretty mixed, I would say.

So, uh, lots of presence from.

, uh, strong presence from Palantir . And also we have an engineer who, uh,

went along through the journey to take, uh, dual lingo through I p O.

So growing from startup to a more established company.

And they have also had, so data experiences, infras

experiences back at Uber.

So a very interesting mixture of engineers with data and other info backgrounds and

across different stages of companies.

Sounds like

a, an all-star team you've accrued so far.

So kudos to, uh, to the, the current success and, um, what would you say is

something that folks can get excited about if joining Blue Sky today?

You know, what's something that you're really excited about going into 2023?

Yeah.

Thank you Tim.

So, really excited about the team and thankful for the opportunity

to work with everyone on a daily basis in terms of what's exciting.

First of all, we are looking for people, what we call our who are merc,

uh, missionaries, not mercenaries.

So first and foremost, they need to, uh, be passionate about our mission of

building next generation cloud infra for data and inner future for ml.

So starting with making data workloads cost efficient and.

Along that journey, we want to also understand their personal

motivations for joining Blue Sky.

No financial reason can be a good one.

Uh, but usually we think it should be the third reason.

Maybe the second, but not the first reason.

But just to understand the personal motivation and where they want to grow.

And if the aspiration is aligned with Blue Sky and also personal

chemistry, founding is very important.

Uh, so you know this.

There will be ups and downs joining an early stage company.

There is bound to be surprises, temporary setbacks, and if, uh,

there are mercenaries, they might be, uh, kind of exiting at the first

sign of things, not going well.

But we need people who would have enough of a conviction, enough

of, of a confidence and grit to go through the challenges.

And I would say also loyalty, uh, with each other.

You gotta

trademark that.

Missionaries versus mercenaries.

That is a fantastic, uh, philosophy.

I think that's, uh, that's really neat.

It's the first time I've, I've heard that before, but I think it, Sam, fantastic.

. It just, it, it makes a lot of sense from a startup environment, you know,

joining for those right reasons and you know, very easily you can see

folks, you know, getting distracted.

You know, especially maybe coming from a FANG environment, right?

Where the, maybe that comp isn't as, as equitable as it might have been in,

in the big tech space, but you know, what is your priorities right now?

What is it that you're really looking for?

Um, I want to close out, uh, our last, uh, few minutes here with a, uh, a fun segment

that we call the Five Second Scramble.

Uh, and in this segment, you know, what I'll do is ask a few rapid fire

questions with you, uh, and you try to sum up your answers within five seconds.

Does that sound, sound?

Sounds good.

Cool.

All right.

Um, so we're gonna start with a little bit of the, uh, some, some business

discussion, uh, business questions, and then we'll, we'll float into some

personal, um, explain your product to me as if I were a five year old.

You want to make the things you do, uh, fast, but also cheap.

Nice.

What problems are you.

We, we help the world finish the transition of their data workloads

to the cloud by making cost efficiency a primary concern, and, uh, keeping

it high with our continuous software based monitoring and optimization.

Who are your users?

Our current target users are, The data infras engineers and

the head of data who manage their company's snowflake instances.

What is your favorite

aspect about working at Blue Sky?

The sense of

being empowered to pursue a direction that we deeply resonate

with and think of meaningful and, uh, work with a group of great profess.

What aspect of your culture do you most fear losing with growth,

being scrappy, and being able to work in the flat way and moving fast

and not having to coordinate excessively.

What about your work keeps you up at night?

It keeps me up at night in a good way because I'm excited about the

future and there are so many things that we need to tackle, uh, but in a bad way

where this is gonna be a long journey.

So I gotta also make sure it's, uh, a marathon, another sprint.

So that's something we, you know, focus on a personal, uh,

physical and mental wellbeing.

So you

all are impacting a lot of verticals within ai.

Um, you know, what about ai?

The future of AI excites you the

most.

What's topic these days are large language models.

They are driven by having high quality data.

One of the reasons why we started Blue Sky from the data angle is a

lot of ML companies first need better data pipelines and better data.

That's a good starting point to contribute to ai.

What about the

future of AI scares you?

What's immediately on top of my mind is people set the proper

expectations of the performance and power of ai, for example, don't blindly

trust it if it's just making things up.

Hmm.

Good answer.

What is one of your favorite.

In addition to thinking about blue sky and innovating, I

enjoy taking time off through hiking.

That's my personal form of meditation.

What do you love most about yourself?

Being driven and even obsessed.

Do you believe that there is life on other planets?

I don't have reasons to dismiss that emotionally.

I hope that is true.

, Tim: what is the favorite, your

Oh, wow.

YouTube.

I watch YouTube a lot.

And what

is your favorite superhero?

um, professor X sounds pretty.

Good Xmen reference.

Good stuff.

Well, I think you have, uh, navigated all of those questions.

Um, and you know, a quick plug as well on, on the YouTube front, we did pick up that

there is a podcast that, that blue sky, uh, uh, host, um, called Above the Clouds.

Do you wanna quickly give a, a shout out as to what it's about?

Yeah.

Thank you,

Tim.

So above the Clouds is our, uh, blue sky production in.

Where we use some site bandwidth to in, to interview the industry, leads in the

data area, uh, and also entrepreneurs, uh, professionals to learn about their

perspectives of the current challenges and the techniques in solving these

challenges like making, uh, improving cost efficiency, uh, et cetera.

So this is our way of, uh, contributing back to the c.

And also, uh, helping integrate Blue Sky into this larger community

of the modern data stack in the cloud-based data and ml, uh, community.

That's

great.

Yeah.

It seems like there will be some crossover between our audience and

yours, so we'll be sure to, to plug that in the show notes as well.

But I just wanted to.

To thank you for spending time with us, uh, Ming Shang.

We are excited for, you know, the future of what you all are building.

Uh, we love the space and I think it's important to have a company

like Blue Sky around to help facilitate the growth of, you know,

other industries like AI at large.

So we are.

We are rooting for you all.

Um, and, uh, thanks again for, for spending time with us.

Thank you very much, Tim.

It's really great to be here and look forward to continuing out that dialogue.

Are you a startup founder or tech leader looking to grow your

engineering or product teams?

If so, hatch, it could be a partner worth exploring.

We've helped hundreds of startups scale their tech teams with relational and

marketing driven recruiting solutions.

Check out Hatch it.io/hire to learn more about how we can help your teams grow.