Harry Mellsop: Kiwi entrepreneur on his new AI-based startup Antioch

New Zealander Harry Mellsop has raised US$4.3 million for his start-up, Antioch, which he says brings artificial intelligence into the physical world.

He explained that Antioch’s tech will allow companies to build new or better AI robots by letting them test new features in software simulations.

Mellsop says this technology will help develop the autonomous driving space, drone technology, the manufacturing sector and several other key industries. 

"I think something that's personally surprised me is a lot of interest from smart security companies, so smart home sensors, to smart doorbells, those sorts of applications as well."

LISTEN ABOVE

See omnystudio.com/listener for privacy information.

Episode Transcript

Speaker 1 (00:06):
You're listening to the Sunday Session podcast with Francesca Rudkin
from News talksb Right.

Speaker 2 (00:13):
This week, expat Kiwi entrepreneur Harry Melsop has raised four
point three million US dollars for his AI startup, Antioch.
The former Tesla engineer and Stanford University graduate, is part
of a team bringing AI into the physical world. So
Antioch has built a software program that lets companies building
robots test them in real world scenarios. So Harry is

(00:34):
just right back in New Zealand. He joins us to
talk through what he is hoping to achieve. Harry, good morning,
Good morning, How are you good? Thank you hey, talk
me through what you've created here in Layman's terms.

Speaker 3 (00:46):
Sure thing, Yeah, So, I mean, you know, at Antioch,
our mission ultimately is to help companies who are building
autonomous systems, whether that's a self driving car or a
robot in a factory or a drone, to sort of
shift all of their validation workflows from hardware into pure software.

(01:07):
And so you know, as you can imagine, if you're
building a self driving car, super important to test that
system thoroughly. And so what that means is, you know,
hundreds of thousands or millions of kilometers driven on real roads,
super expensive, you know, the car's breakdown, and so we're
really helping all of those companies move their workflows into

(01:29):
what's basically you can think of it as a really
high fidelity video game engine that really closely models the
real world.

Speaker 2 (01:37):
Okay, so what other kind of industries would be using
this technology?

Speaker 3 (01:42):
Yeah, so we work with Yeah, so you know, beyond
the autonomous driving space, beyond the drone space, we're seeing
a lot of interest from manufacturing, from logistics. I think
something that's also personally surprised me is a lot of
interest from smart security companies. So you know, smart home sensors,

(02:04):
smart doorbells, those sorts of applications as well, because you know,
you can imagine for them at testing that that those
devices are properly detecting as they should be is a
high priority too. So there's really quite a wide gamut
of applications there.

Speaker 2 (02:19):
I'm assuming that there's a gap in the AI market here,
is anyone else doing this?

Speaker 3 (02:25):
Yeah, So what we typically see is that companies who
are trying to make this move into software simulation him
to try and build that capability internally and so you know,
for example, when when I finished my master's degree, my
first job out of that was at Tesla's autopilot team.

(02:46):
We had a fantastic simulation team there internally, but that
costs tends to hundreds of millions of dollars to kind
of build out within the company, and you know, we
don't think that that's the way it should be, right.
We want to make sure that we're kind of equalizing
the playing field so that newer and smaller startups can
also kind of participate.

Speaker 2 (03:07):
So that's yeah, so quite different industries using this technology,
do you need to then design it? Are you constantly
designing for different specific companies and industries.

Speaker 3 (03:23):
Yeah, no, that's a great question. And so I think
there's kind of two parts to how we structure the business.
So the first is that we build the simulation technology
in a very bottoms up fashion, so we add support
for you know, maybe it's motives, or maybe it's cameras
or radars or so on and so forth, and then

(03:43):
that kind of allows us to you know, service a
more diverse stet of customers ultimately, because you can start
to think of all of these different categories, whether it's
a drone, or autonomous vehicle as a sort of you know,
bringing together of those different foundational technology.

Speaker 2 (04:00):
So you've kind of got it sitting there and it
depends on what the client's going to need. You kind
of take that and take that and take that and
put it all together.

Speaker 3 (04:06):
Yeah, that's exactly right.

Speaker 2 (04:08):
I'm sure it's not quite out works at all.

Speaker 3 (04:10):
Eric, No, No, that's actually that's pretty accurate.

Speaker 2 (04:14):
How big do you think the physical AI industry is
going to be?

Speaker 3 (04:18):
Yeah, No, that's a great question, and it's something that
we have spent a lot of time thinking about before
we started this company. I believe that it's going to
be the largest industry in the world over the next
ten to fifteen years. I think that we you know,
we've seen with the rise of large language models and
companies like open ai, that the world of desk work

(04:43):
has fundamentally changed forever. And I think we're going to
see those same trends playing out in the physical world
as well. And ultimately those industries are really what matters, right,
you know, we need to we need to eat food,
you know, we consume things in the physical world, and
so the scope for impact there is enormous.

Speaker 2 (05:02):
So Harry, what is this funding going to do for you?

Speaker 3 (05:06):
Yeah, so we're using it to expand the team out.
We've been building the team primarily in New York City
and so you know, bringing in new engineering staff that
allows us to take on new customers. Currently, we are
in a unfortunate position where we're pretty saturated in terms of,
you know, the number of customers that we can kind

(05:26):
of help, so we we pretty urgently need to expand
a little bit there. And then also, you know, the
other surprising piece for us has been the need to
kind of produce content in these simulations. So you know,
if some of our customers have the need for particular
warehouses to be modeled or you know, home environments or whatever,

(05:46):
that looks like we actually need quite a large treaty
artistry capability as well, so bringing some of those folks
into the team.

Speaker 2 (05:56):
This isn't your first company. You also founded a company
called Trendspose, which you sold after setting up is anti
K a project that you're really keen to kind of
get up and running and see where it will go
and stick around whether or are you sort of planning
on getting it up and running and selling that as well.

Speaker 3 (06:14):
Yeah, it's a good question. Now, I think for us,
we kind of see the opportunity here as being enormous.
You know, I think we're all very aligned on this
idea that this physical AI industry is going to be
world defining over the next ten to fifteen years, and
we think that we're building a product which is, you know,

(06:35):
really very necessary. Everyone who's participating in that space is
going to need a solution like Antioch. So for us,
this is a business that we're going to want to
hold on to and we hope to grow.

Speaker 2 (06:47):
Harry, You're most welcome not to answer this question, but
can I ask how old you are?

Speaker 3 (06:52):
I'm twenty seven, cry.

Speaker 2 (06:54):
Key, Okay, this is your second business, yes, yes, yes, okay.
Oh look, Harry, really impressive stuff. Really keen to keep
it on it and see how it goes. Thank you
so much for your time, have a lovely Christmas and
enjoy your time back in New Zealand.

Speaker 1 (07:10):
For more from the Sunday session with Francesca Rudkin, listen
live to News Talks a B from nine am Sunday,
or follow the podcast on iHeartRadio.