0:00

What are the challenges that you see and, and that

0:02

you guys are trying to tackle?

0:03

Oh man. Let's see. The

0:05

market's insatiable. The customer

0:08

base wants more, wants

0:10

it faster, wants it in different

0:12

flavors, wants it in different ways

0:15

packaged differently sustainably, of course.

0:17

So that's really

0:20

hard as well to try

0:22

to get more from

0:25

the existing team, sometimes even

0:27

a smaller team because of economic

0:29

issues. So the old

0:31

adage of how do we do more with less

0:33

is, is very, very present

0:36

right now. So that's, that's a

0:38

big challenge. The, the other

0:40

challenges are that this stuff is

0:42

really hard. You know, chemical

0:44

development, material science development

0:47

is hard. It's, it's multidimensional.

0:49

It has an unbelievable amount of variables.

1:02

A key component of the modern world economy,

1:04

the chemical industry delivers products and innovations

1:07

to enhance everyday life. It

1:09

is also an industry in transformation

1:12

where chemical executives and workers are delivering

1:14

growth and industry changing advancements

1:16

while responding to pressures from investors, regulators,

1:19

and public opinion, discover how

1:22

leading companies are approaching these challenges

1:24

here on the chemical show. Join

1:26

Victoria Meyer, president of Progressio

1:28

Global and host of the chemical show.

1:31

As she speaks with executives across the

1:33

industry and learns how they are leading

1:35

their companies to grow, transform,

1:38

and push industry boundaries on all

1:40

frontiers. Here's your host,

1:42

Victoria Meyer.

1:46

Hi, this is Victoria Meyer. Welcome

1:48

back to The Chemical Show. Today

1:50

I am speaking with Ned Weintraub,

1:52

who is the Chief Revenue Officer

1:55

at NobleAI. Noble's a

1:57

pioneer in science-based

1:59

AI solutions for chemicals and

2:01

material companies. And Ned has been

2:03

involved in digital innovation

2:05

for industry throughout his career at

2:08

companies including Seven Signal

2:10

Verana and HP Cloud. We're

2:12

gonna be talking about business challenges

2:15

with accelerating chemical development,

2:17

innovation, how AI fits into

2:19

that and a whole lot more. Ned,

2:21

welcome to The Chemical Show.

2:23

Thank you for having me, Victoria.

2:25

So you've spent your career in

2:27

technology and growth oriented firms.

2:29

What got you started in this space

2:31

and what ultimately brought you to NobleAI?

2:34

Yeah, it's a good question. For me,

2:37

I've always tried to

2:39

work for Mission-driven companies, and

2:42

Noble is a mission-driven company. We

2:44

believe that AI

2:46

can solve the world's problems

2:49

better than it can destroy

2:51

the world. And so we believe that

2:53

that. This is, this

2:55

is a fantastic way of, of really

2:57

trying to solve health

2:59

issues, environmental issues all

3:02

sorts of different ways that we can,

3:04

you know, get over what we're trying

3:07

to, to fight every day.

3:09

Yeah. Makes sense. And I, and I like that mission driven

3:12

space because I think we sometimes lose

3:14

track of that in business. And yet

3:16

it all starts out with a, a bigger purpose.

3:19

You got it. That's exactly right. It helps us

3:21

get up in the morning.

3:23

Absolutely. So the chemical industry

3:25

it is definitely in a period of accelerated

3:28

innovation, right? We're seeing this,

3:30

in fact, you and I recently met up at

3:32

ACI, One of their themes was innovation.

3:35

So innovation is everywhere,

3:37

and yet we're still challenged

3:40

in commercializing these new ideas,

3:43

new products, new innovations. So

3:46

first of all, what trends do you really

3:48

see driving innovation in the chemical

3:50

space when you go out and talk with customers?

3:52

I would say one

3:54

of the biggest ones, was sustainability.

3:57

Sustainability has finally arrived

3:59

and moved from sustainability

4:02

washing to really

4:04

putting budgets behind it.

4:06

Again, a mission. Where

4:09

companies can feel good about

4:11

what they're doing and really driving

4:14

sustainability through everything.

4:16

And frankly, product development

4:18

and, and optimization is a big one. That's

4:21

the biggest one. But I would also say

4:24

adapting and agility to regulatory

4:26

issues. That was the other big focus

4:29

and we hear this really from every single

4:31

one of our customers, which is the

4:33

world is changing fast. How

4:36

do we adapt? How do we become more agile?

4:38

How do we become proactive instead

4:41

of reactive to a lot of these things?

4:43

Those are the two biggest ones that we see

4:45

kinda driving a lot of the

4:47

the innovation.

4:49

Yeah, so I think the drivers are strong

4:52

and as you say, they are mission-based in many

4:54

ways, and yet there are still

4:56

some challenges. What are the challenges that you see

4:58

and, and that you guys are trying to tackle?

5:00

Oh man. Let's see. The

5:03

market's insatiable. The customer

5:05

base wants more, wants

5:07

it faster, wants it in different

5:09

flavors, wants it in different ways

5:12

packaged differently sustainably, of course.

5:15

So that's really

5:17

hard as well to try

5:19

to get more from

5:22

the existing team, sometimes even

5:24

a smaller team because of economic

5:26

issues. So the old

5:29

adage of how do we do more with less

5:31

is, is very, very present

5:33

right now. So that's, that's a

5:35

big challenge. The, the other

5:37

challenges are that this stuff is

5:39

really hard. You know, chemical

5:42

development, material science development

5:44

is hard. It's, it's multidimensional.

5:47

It has an unbelievable amount of variables.

5:49

Everything from price to market

5:52

pressures to supply chain

5:54

risk, personnel risk. This stuff

5:56

is really hard. So how

5:58

can technology

6:00

like science-based AI help

6:03

in these situations, and that's really

6:05

why we're here.

6:06

And it seems to me, I mean, I think part of it,

6:08

you talk about the people challenges and

6:10

just the fact that it's really hard. I

6:12

know what we're seeing is the continual

6:15

graying of the chemical industry.

6:18

Although some of us like better living through chemistry,

6:20

so we wash that gray away and then

6:22

what have you, right.

6:24

Absolutely.

6:25

But I think what I'm hearing from people

6:27

across the industry is a real concern

6:30

that they're losing really

6:32

experienced staff. Yeah,

6:34

across the board, and it seems particularly

6:37

in the product development

6:39

and formulation development space because

6:42

they're retiring, right? So they're just, they're,

6:45

they're moving on to greener pastures

6:47

or however we wanna say it. Um,

6:49

and, you know, we, we just don't have the same

6:52

knowledge base in the industry. And

6:54

it seems like, AI is

6:56

one of the ways that we can really

6:58

harness and leverage some of

7:00

that existing knowledge,

7:02

even perhaps when the people that

7:05

developed it have moved on.

7:07

Yeah. Unbelievable. I thought

7:10

there's two issues. Number one, you're absolutely

7:13

right. Senior leaders

7:15

are retiring, right? They've been in

7:17

this industry for a very long time. But the

7:19

other one, which is really

7:21

interesting is that there's a talent

7:24

war. And people

7:26

are leaving. It's not the old days of

7:28

sticking with your company and getting a gold

7:31

watch at the end of 30 years. The

7:33

world has changed. So how

7:35

can companies de-risk? You

7:38

know, the personnel flight, the brain drain

7:41

as they call it. So is there a way to

7:43

really mimic or really

7:45

institutionalize or codify

7:47

that institutional knowledge and

7:50

not be at risk when that person walks out

7:52

the door. That's also a really interesting

7:54

conversation that we're having with customers. So,

7:57

you're a hundred percent right there.

7:58

Right. So I mean, you guys are at the, the

8:00

front end of moving

8:03

AI into product development and developing

8:05

solutions to support the industry. What

8:07

do you see as the role and what are the conversations

8:10

that you're having about how

8:12

people want to integrate AI in this

8:14

space?

8:16

Yeah. The best thing that's happened to

8:18

us is generative AI, right?

8:20

You know, the chat GPTs and the Geminis

8:23

and all of these others have really brought

8:25

it to the forefront. Now remember,

8:28

artificial intelligence has been around 30

8:30

years longer than the internet has been

8:32

around. So just just

8:34

to give people an understanding, this is

8:36

not, it's 15 minutes of fame.

8:39

It's been around for a long time. It just

8:41

now has, because of the internet,

8:44

has been able to take all of this

8:46

data, the oceans of data

8:48

that they can mine. What's

8:51

great for creating speeches

8:53

and my kids' homework doesn't

8:55

work for science necessarily

8:57

because there's not a lot

8:59

of data. There's not an ocean's amount of data

9:02

because we're trying to create

9:04

things. And so, you know,

9:06

if it were that easy, everybody would do it. We

9:09

have to be able to help

9:11

our customers drive

9:13

new innovation optimize

9:15

current products with not

9:17

a lot of data or completely

9:20

spread out data. For us, this

9:22

notion of specialized

9:24

AI, or science-based AI,

9:27

especially in chemical and

9:29

material science is really

9:31

critical because we can do a lot with

9:33

very little data. We can be very

9:35

prescriptive in what problems we

9:38

we're being asked to solve and

9:40

really be able to accelerate based

9:42

on that. That's a big piece

9:45

of, it's great that we have generative AI

9:47

but you know, specifically focused

9:50

in this world that you and I live in,

9:52

it has to be specialized.

9:54

Right. Yeah. It's interesting. So you say that

9:56

there's not a lot of data. I think people would assume

9:58

we are awash with data. And certainly

10:01

it's true in, when I think

10:03

about chemical companies and

10:05

just their overall business and business

10:08

operations, we have a ton of data about

10:10

customers, we have a ton of data ton of data about

10:12

manufacturing we probably

10:14

have a lot of it about product and product development.

10:17

Although I will say, I reflect back

10:19

and think about my time in industry,

10:21

when I worked really closely with our

10:23

formulation guys to help get new products out into

10:26

market and stuff, I was sometimes

10:28

shocked by like how many data points,

10:30

how few data points were actually on a curve

10:32

or on a graph or what have you. So it's

10:34

it's interesting twist on this

10:36

'cause I sometimes think. We

10:39

feel like there's just all this data

10:41

and yet maybe it's not always the right data

10:43

or in the right place at

10:46

the right time.

10:47

All of it. Absolutely. All, all

10:49

of the above. You're right, these companies have

10:51

been around for. You know, some a

10:53

hundred years and you would think that

10:55

the data is A, there

10:57

and B accessible. And

10:59

very often it's not.

11:01

And neither one or one of them. And

11:04

it's, it's spread all over the world in

11:06

different lab notebooks, physical and electronic,

11:08

maybe electronic. It's

11:11

very, very difficult. And so, how

11:14

do we help customers get started without

11:17

having to rake the ocean full

11:19

of ones and zeros to get started? That's

11:22

really where building that

11:24

science into the AI

11:27

right from the get go alleviates

11:29

a lot of that need for a lot

11:31

of that data. Because we already know a lot

11:33

of that information. We can build it in institutionally

11:37

and really accelerate that development.

11:40

So what does that look like? So say more a little bit about

11:43

that. 'cause I think it stills kinda,

11:45

you know, 20,000 miles up in the, the

11:47

atmosphere. You know, what

11:50

does it mean to be science-based? Is it generative? this generative?

11:53

And how does it work?

11:56

Yeah. So it's,

11:58

it is generative in a way

12:00

because we can generate new

12:03

insights, but fundamentally,

12:05

when we talk about specialized AI

12:07

or science-based AI, we're

12:09

building the fundamentals of

12:12

physics and chemistry into

12:15

the models that we build with our customers

12:18

in partnership. I

12:20

always like to say for the senior executives

12:22

who aren't necessarily scientists, and I don't

12:24

have a scientific background, you

12:27

know, elephants don't fly. We all know

12:29

elephants don't fly. But with,

12:31

with commercial AI,

12:33

something you get off the shelf, you

12:35

have to train it, that elephants don't fly

12:38

and therefore. It takes time.

12:41

And then a lot of the answers early on

12:43

you get, well wait a second, this doesn't make any sense.

12:45

Why are we even going down this path? So

12:47

they give up. Where by building

12:50

all of that knowledge upfront into

12:52

the models and then using different

12:54

models, solving

12:56

different problems that

12:59

really accelerates the insights

13:01

and that gets us to, even in the first

13:03

rudimentary models that we build with our

13:06

partners. There

13:08

are aha moments. We've, we've solved

13:10

some very fundamental problems

13:13

for customers that have been struggling with,

13:15

you know, maybe it's a PFAS chemical

13:17

that they're trying to get out of one of

13:19

their formulations. We were able

13:21

to, you know, give them insights within

13:24

30 days, something that they've been trying

13:26

for years to solve, or at least, you

13:28

know, the last two years, within 30 days,

13:31

we gave them directionally approaches

13:33

to head to. So by building

13:35

that institutional and, and that scientific

13:38

knowledge into those models early

13:40

on, we really

13:42

accelerate that. And then we can train

13:44

those models as we continue. And

13:47

then customers use them once they're mature

13:50

to drive insights to really

13:53

do a lot of testing that they wouldn't

13:55

otherwise be able to do on

13:57

a bench, if you will.

13:59

yeah. Makes sense. Yeah. Yeah.

14:01

30 days seems fast. I know. In the world of chat,

14:03

GPT, if I have to wait longer than 30 seconds

14:06

for my answer, it seems like it's taking a

14:08

long time. But, but to your point, this

14:10

is I guess a much more rigorous approach

14:13

as needed, right? I mean, it has to be rooted

14:15

in the scientific principles, whether it be chemistry

14:18

or physics or material science to, to

14:20

make that happen.

14:21

right. We're, we're dealing with scientists, by the

14:23

way, so, you know, they, they want facts.

14:25

They

14:26

They want facts and, and they're probably

14:28

a little bit risk averse. So let's,

14:30

let's talk about those risks. So what,

14:32

you know, I think what are the risks that

14:34

you guys see when, or that you

14:36

talk about and you work to alleviate

14:38

with your clients when you think about using AI

14:41

and product development?

14:42

Yeah. So honestly,

14:45

the risk that we see our people

14:48

are gripped by this notion of not having

14:50

all the data in one place. And

14:52

honestly it's a bit of the

14:54

boogeyman that, you know, part

14:56

of the industry who is about

14:59

trying to collect all of your data in one place

15:01

before you get started. That's their

15:03

message. And the reality of it is, is

15:06

that the risk of not

15:08

getting started now. You're

15:10

allowing your competitors to, to distance

15:13

themselves from you, right? To either gain the

15:15

edge or to expand if, if you can't

15:17

do that, you know, part of the reason why

15:19

we do a lot of work with mid-size companies

15:22

is because they can't throw

15:24

money and bodies at this internally,

15:27

and they have to do what

15:30

they have to do in terms of closing

15:33

the gap with those big companies. And

15:35

we see that every day and they genuinely

15:37

see AI as.

15:40

Both a panacea, so we have to kind of temper

15:43

their enthusiasm, but also give

15:45

them the true value, you

15:47

know, vision of what it can really do

15:49

for them. So the risks, getting back to your

15:52

question, are just getting started.

15:54

That's one. The other risk

15:56

is the industry has spent

15:59

a hundred years hugging their

16:01

IP and not allowing it

16:04

out into the world because genuinely

16:06

that is the keys to their kingdom. How

16:08

do they work with partners,

16:11

feel comfortable about working with partners,

16:13

but still have the.

16:16

Security literal and figurative,

16:19

figuratively, to be able to

16:22

really collaborate with people outside

16:24

their four walls. And that's a, that's

16:26

a perceived risk as well, right? The

16:29

cloud industry had to go through this,

16:31

right? It's all of our data needs to be

16:33

here. And then people realize that

16:35

AWS and Azure are probably

16:37

even more secure than your own network

16:39

itself. So, these are early

16:41

days. Those are the, those are the

16:43

challenges that we work through with our customers.

16:47

I can see that. And certainly the ip, the

16:49

intellectual property and data privacy

16:52

is, is probably the thing I hear the most.

16:54

Um, in many ways it's maybe the most misunderstood

16:58

in my opinion. And, and I've done some work

17:00

around this some folks that. You

17:03

know, there's this perception of, oh, if I, if

17:05

I put it out there, it's there for the public domain,

17:08

Right, It's like, well, no, no, no, there are firewalls.

17:10

And by the way, don't put your don't

17:12

put your test data into chat GPT

17:15

because chat GPT is

17:17

open domain, right? So buyer

17:19

beware. But there are other, I mean, heck,

17:21

there's a version of chat that you can buy that's private and

17:23

obviously when, if you are working with a company like

17:26

Noble, there's firewalls and

17:28

privacy protections to protect. All

17:30

that data.

17:31

Yeah. I'll, I'll even go one further. So

17:33

you're a hundred percent right. There's

17:36

the risk on the generative AI side that

17:38

you are putting all of your. Your

17:40

information out on the internet. So

17:43

our customers do have and

17:45

are working through these policies

17:47

for their employees. So that should be,

17:50

that should be looked at where

17:53

science-based AI folks

17:56

like NobleAI work.

17:58

Is within their customers

18:01

domain. So we have the ability

18:03

to build these models

18:06

and serve them to our customers

18:08

within their private cloud.

18:11

So that's a big differentiator for

18:14

us because we feel like, yes,

18:17

we're not gonna try to change the hearts

18:19

and minds about people's ip,

18:21

right? It's, we don't have enough time. To,

18:23

to try to, to create a sea change

18:26

there. So for us, we feel like

18:29

we, we have the ability to work within

18:31

that. You know, that framework,

18:33

and that's been very successful. The second

18:35

thing is, the other challenge

18:37

is in the AI world

18:40

is this notion of, of who

18:42

owns the models, right? Who owns

18:45

this ip, right? Is it

18:47

the, is it the AI company

18:49

or is it the chemical company that's

18:51

bringing that data? And historically,

18:54

all of the last five to 10

18:56

years. The AI companies

18:59

have said, oh, no, no. Those are our models.

19:01

Those are our models. And so it's really

19:04

set up to be this very confrontational

19:07

you know, is it ours? Is it theirs, is it co-owned?

19:10

How do we do that? If we wanna publish

19:12

it, I mean, it, it can become a nightmare.

19:15

We've taken a different approach. We,

19:17

we build customized

19:20

models specific to our customers,

19:23

and they own those models because

19:25

for us, it's, it's important

19:27

that they can build

19:30

from there. It's theirs. I

19:32

always use the, the analogy

19:35

of. Steven Spielberg

19:37

writes a screenplay.

19:39

He wants a movie made. He goes and raises

19:42

money and go, gets it made. He takes it to

19:44

a filmmaker. He takes it to, you

19:46

know Skywalker Rancher and

19:49

they own the way the movie

19:52

gets made, the special effects

19:55

and all of the different ways that it becomes

19:57

a fabulous Steven Spielberg movie.

20:00

Steven Spielberg owns that movie. The

20:02

movie company doesn't own that movie.

20:04

And so we, that's our

20:06

approach. We feel as though, and our customers

20:09

appreciate that because they

20:11

don't have to focus on. Oh

20:13

my God, are they gonna turn around and sell this

20:15

to a competitor? Which is obviously

20:18

in business a very real situation.

20:20

So that's, that's how this

20:22

industry is starting to evolve. We feel like

20:24

we're on the forefront of it.

20:26

Yeah, that ownership risk, that's great.

20:28

The other thing. That I hear,

20:30

and this is a widespread concern

20:32

regarding all AI

20:35

and all generative AI, is

20:37

that we're using a limited data set

20:39

and that we're just kind of creating this very

20:41

narrow spiral based

20:45

on limited data. And once

20:47

it gets skewed, the truth gets

20:49

skewed.

20:51

Yeah. The bias, the bias comes in.

20:53

Yeah. That bias is always

20:56

a, an omnipresent

20:58

thought around building

21:01

these models in AI, it's

21:03

the, it is one of the biggest reasons

21:05

why companies should be partnering

21:07

with companies like NobleAI

21:10

because. Institutional

21:13

bias happens within the same four walls.

21:15

It's the same people building these models,

21:17

and they, and they own

21:19

that box. Now, AI

21:22

does a great job of, especially

21:24

science-based AI and specialized

21:26

AI to broaden

21:29

your horizons, right, your design space,

21:32

but. Bringing

21:34

in people with external experiences,

21:38

a from either people within the industry

21:40

or even better yet, people outside the industry.

21:42

What are the people in oil and gas doing?

21:45

Although it's related, upstream

21:47

is very different. What, how, what are they,

21:49

what are they doing in exploration? What are people

21:51

doing for alternative energy? Is there something

21:53

there that we can bring to the packaging industry?

21:56

What are people doing in, in Biosynthetics

21:58

and, and what can we bring to that? So,

22:01

you're right, that's if you are

22:03

trying to do it internally

22:06

and without multiple

22:08

different ways of building these models

22:11

with NobleAI. I think for our,

22:14

our domain of different models, we've got

22:16

over 45 different ways

22:19

of building models that

22:22

could all be combined. It's not just

22:24

kind of repurposing the same model over

22:27

and over. 'cause that generates bias.

22:29

Yeah. Yeah. And I suppose once

22:31

a solution is identified,

22:34

let's just say you, you brought up the PFAS example.

22:36

Once the new alternative formulation

22:38

that replaces PFAS is identified,

22:40

there's still lab work that gets done. There's

22:42

all kinds of testing. And so new data

22:45

is created

22:47

absolutely.

22:48

into the model. As long as I guess it, you

22:50

know, you understand where it goes in and to

22:52

your point, there's, there's always biases always

22:54

existed in,

22:55

Always existed even more so

22:57

without

22:58

let's just say it's in

22:59

well, yeah. I mean, even before, right?

23:01

I mean, you have your scientists who are brilliant,

23:03

but they know what they know. That automatically

23:06

instills bias. But you

23:08

brought up a very good point which is

23:11

this notion that AI is

23:13

going to wash away

23:16

jobs. That may be

23:18

the case in other industries, but

23:20

absolutely not in our space.

23:23

The need for, first of all, these

23:25

are scientists. They don't trust. Anything

23:27

they've gotta verify, they've gotta double verify,

23:29

they gotta triple verify. So whatever

23:32

we do in silico on the computer

23:34

is going to be wet verified.

23:37

It's gotta be verified in a lab

23:39

that, yes, this makes sense, I'm

23:41

replicating this and therefore

23:44

this gets me to my

23:46

ultimate goal faster. So we

23:48

are not seeing that at all. What we see

23:50

is the advent of

23:53

doing. A lot

23:55

more, a lot faster.

23:58

They have moonshot projects that have been

24:00

on our whiteboard for two years and

24:02

haven't moved. And you know, there's a sign

24:04

that says Do not erase. And you know,

24:07

people have left that, but it's always stayed in the upper

24:09

left hand corner. Now these

24:11

things are starting to get pulled into view. These

24:13

same folks, they're not losing their jobs

24:16

by any stretch. They now get to work

24:18

on. Four times

24:20

the amount of projects than they ever had,

24:22

so that's been very exciting.

24:24

That's cool. That's very cool. So, so

24:26

this is maybe a good segue to our next topic,

24:29

which is really around customers

24:31

and customer acceptance, maybe even the customer

24:33

experience. And I know that Ned, you're an expert

24:36

in sales and business development and that's

24:38

the role that you've played, um, with

24:40

a number of companies really helping drive

24:42

that customer and that value. Um,

24:45

and I know that you're out talking to chemical

24:47

companies and people across the value chain

24:50

every day. What are your

24:52

customers excited and

24:54

and or concerned about when

24:56

they think about bringing in

24:58

an AI based solution to their company?

25:01

I would say the, the folks

25:03

who are looking inside the operational

25:06

folks are worried about disruption,

25:09

right? Transformation is scary. And

25:11

so that, that's

25:14

number one. Number

25:16

two is do

25:19

we really have the people

25:21

in house that can leverage it?

25:24

It's not just enough for a partner to

25:26

hand this to us and run with it. We

25:28

have to have the people that can run with it. And

25:30

very often there is

25:32

some change over there. But

25:35

I would say for the most part

25:38

ultimately because it is new

25:41

the finance folks can't

25:43

really qualify it right? Or

25:45

quantify it actually. Therefore,

25:48

it becomes this, are we risk averse

25:51

are we really ready for this? So

25:54

my job as a

25:56

business development person and my team

25:58

is really there to help

26:01

them understand what the business value

26:03

is, right? The scientific value, I think

26:05

is pretty demonstrable. The

26:08

economic value is

26:10

really where the senior

26:12

executives want to be able

26:14

to sign off on it, but they're

26:16

not necessarily willing to jump into

26:18

the deep end of the pool without some,

26:21

Either a reference or, you know, some business

26:23

case built. So we spent a lot of time,

26:26

you know, what, what would an acceleration

26:28

of this project or. Financially

26:31

for you what are the risks that you're seeing

26:33

now from a supply chain? We have one customer

26:36

who had to take a product off the market for eight

26:38

weeks because one of their small little chemicals.

26:41

Yeah. Eight weeks is a major

26:42

That's a lot. Yeah.

26:43

a lot. I mean, so it's millions of dollars.

26:46

And so when you have that

26:48

and it's visceral like that is.

26:51

You, you get to figure that out pretty quickly. But

26:54

there are others that are just

26:56

trying to figure this stuff out. business

26:59

wise, it has to move a needle,

27:01

right? We always talk about it's gotta save

27:03

money, it's gotta make money, it's gotta de-risk,

27:06

or it's gotta transform. If you can't

27:08

do two of the four, then,

27:11

you know, probably shouldn't do it.

27:13

Yeah, action then.

27:14

that exactly right.

27:16

Who usually brings you in? Where does that

27:18

happen? Does that happen at. The,

27:21

you know, at the the lab level, let's just

27:23

say, or the product development guys, is it, is

27:26

it the executive team that says, oh yeah, we know we need

27:28

to do something different. Where do you see,

27:30

how do you guys normally enter an organization?

27:33

And then, we kind of touched on

27:35

this, there's obviously different organizational

27:38

priorities depending on where you sit and what

27:40

you're looking at. How do you bridge those

27:42

gaps?

27:44

Yeah, we just met with the CEO

27:46

of a very, very large Fortune

27:48

1000, maybe even 500

27:50

CEO and his entire executive

27:52

team, and they broke it out into four stages

27:55

and research and development for a chemical

27:57

company has a very tall

27:59

pole in that tent, so is manufacturing

28:02

and engineering. So they

28:05

bucket their priorities

28:07

based on. Revenue.

28:11

Right? I mean, that's ultimately, especially

28:13

if you're a publicly traded company, it's,

28:15

it's, it's what moves the needle.

28:18

So who brings us in?

28:20

To get back to your question, number one

28:22

is very often it

28:24

will be a product development.

28:27

I. Manager, someone who

28:29

is either behind the

28:31

eight ball on their product development

28:34

goals, right? Their product is delayed,

28:36

it's over budget. Those are the folks

28:38

who have the budget. But they're

28:40

not the ones who can go and run these

28:42

experiments. They're the ones who then

28:44

have to bring us into

28:47

the data science teams or the r and d teams

28:49

very often. We need all three of those

28:51

to get consensus. It's a challenge

28:54

in my world of, of selling and,

28:56

and business development because

28:58

you do cross all of these domains, right?

29:00

If you're selling IT security, you get to

29:02

sell to the security team and it's

29:04

a pretty linear path for

29:06

us. It cuts across all

29:08

business lines. It cuts across manufacturing,

29:12

it cuts across engineering, from how do we get

29:14

from the bench to the market? either

29:16

the r and d team brings us to product development.

29:18

Product development brings us to the r and d

29:21

teams, the two. But nothing

29:23

happens until we're speaking with senior

29:25

executives because they're the folks who are looking

29:27

outside the boat, as we say. And

29:30

they're looking for icebergs. They're

29:32

the ones who are saying, how can we

29:34

get more gas into this engine?

29:36

How can we do it without, without hiring

29:39

a boatload of people without spending

29:41

unbelievable amounts. Spending millions

29:44

to make millions doesn't make sense. And so

29:46

they're the folks who ultimately we

29:48

need to get to, to really drive this.

29:51

Yeah, makes sense. And I mean, ultimately

29:53

they hold the purse strings and make the big decisions.

29:56

It also strikes me, Ned, that there is there's

29:58

a real need for change management because

30:00

this is a change in

30:04

business processes, um,

30:06

that have probably been

30:08

in place, as you say, you know, maybe for a hundred

30:10

years in some cases. And it's a change

30:13

that hits, interestingly, not just

30:15

r and d and product development,

30:18

but it's also a marketing effort and it's potentially

30:21

a manufacturing and engineering effort.

30:23

How do you see this playing out?

30:25

The, the change management of

30:28

introducing really

30:30

a significant new tool and, and new approach

30:33

to chemical innovation

30:35

in a company.

30:36

Yeah, it's really interesting.

30:39

Some of this is just absolutely

30:41

institutional. So part of what

30:43

we work through is are they

30:46

culturally ready to really

30:48

adopt a change. Now we do

30:50

it. In a crawl, walk, run

30:53

methodology. So we're not asking people

30:55

to burn the boats and the bridges and adopt

30:57

this new path. So we, we,

30:59

we bring them along on this journey,

31:02

but you're right, it's everything

31:04

from administrative. How do we deal

31:07

with people getting access to

31:09

our systems? How do we give

31:11

them access to our data? Do

31:13

we want to, who, how do we minimize

31:16

that scope? All sorts

31:18

of different processes. That's

31:20

why this crawl, walk, run definitely

31:23

works because they get the taste

31:25

of it, they see the value of doing

31:27

one or two projects, and then

31:29

our customers almost across the

31:31

board, add multiple use cases

31:33

and models onto the platform. So

31:36

you're a hundred percent right, especially

31:38

in new technology. Those

31:41

who are looking out of the boat tend to

31:43

adopt earlier and make it happen,

31:45

right? Transformation is never, is

31:48

never easy, and so it's

31:50

hard. Exactly.

31:52

hard stuff, so that's great. What's, so,

31:54

what's next for you? What should we be looking

31:56

at for NobleAI in 24?

31:58

What should we be looking at for AI and

32:01

chemical innovation, uh, as we look

32:03

ahead into 2024?

32:04

I think you're gonna see the rise

32:06

of more and more partners adopting

32:09

the specialized AI. For

32:12

us it's just, going

32:14

deeper and wider with our customers

32:16

finding more opportunity to show

32:19

Yeah.

32:20

really, Victoria, there's no, there,

32:22

there's nothing that we can't work

32:24

on and move a needle

32:27

on. If it has anything to do with science, we

32:30

wanna, we want to at least take a shot

32:32

at it. The second thing that

32:34

we're really working on is expanding

32:37

our presence through Microsoft. Microsoft

32:39

is one of our lead investors. They say

32:41

tremendous opportunity. Not only

32:43

are they an investor, but they're also a customer.

32:46

They wanna bring this science-based

32:49

AI or, you know, science AI for

32:51

science to their customers.

32:53

And so that's really been a, a

32:55

starting to really take off. We're

32:58

gonna be down at CERA week with them

33:00

and very excited for that

33:02

as well. So

33:07

We've come to the end of today's podcast. We

33:09

hope you enjoyed your time with us and want to learn

33:11

more. Simply visit TheChemicalShow.

33:14

com for additional information and helpful

33:16

resources. Join us again next time

33:19

here on The Chemical Show with Victoria

33:21

Meyer.

33:27

yeah, there's just a ton for us to do,

33:30

but boy, there's, you know, right in our

33:32

sweet spot. There's just a ton

33:34

of customers to work with and ton

33:37

of problems to solve.

33:38

Yeah. Cool. Awesome. Well, Ned,

33:40

this has been great. Thank you for joining us

33:43

today on The Chemical Show.

33:44

Absolutely. I really appreciate you inviting

33:46

me, Victoria.

33:47

Yeah. I'm so glad to have you here and thank you everyone

33:49

for listening. Keep listening, keep following,

33:52

keep sharing, and we will talk again soon.