Driving conversational AI adoption from China to Singapore then the world with WIZ.AI's Robin Li

Thank you very much Robin for coming on

call with us and I'm really interested

to learn more about your experience

coming from many different tech

companies in China and then eventually

coming here to Singapore to work with AI

on what they're building which is not

just a Southeast Asia solution for

conversational AI but really a global

one right so they have clients all

across the globe including even Latin

America quite an interesting company and

always great to hear from their leaders

I think we've had Jennifer

a couple years ago on on the show and

now great to to meet another leader at

Viz.AI. Yeah. Maybe you can do a quick

intro of yourself for our listeners and

for our audience and then we'll go from

there. Yeah.

>> Okay. Thank you Paul. I'm Robin from

China and I work for the AI companies

for about more than 10 years and I work

for the enterprise technology industry

for more than 20 years. Okay. So

actually I know Jen phone and we for a

couple years before and what I know is

that I get the journey about weed that

they did the very well on the voice

board and also have a very strong

customer base in Asia this area this

region and I think that is all wrong

that dragon best dragon festival

>> dragon boat festival this year and James

won't come to Shanghai and discuss with

me about the new vision about we about

the AI partner partner area. So that's

affected effectively in my interesting.

Okay. So we discussed a lot about how

the large language model is changing the

game in the enterprise technology and

also that we exchange a lot of insights

and what we know in the comparing

between the Asia and also the main

China. So that lead to the question that

the general asked me if that is I'm

considering about opportunity in

Singapore. So that's why I'm here

because the to me that is a very natural

decision and very quick decision making.

You consider about the macroeconomics

environment. You consider about the the

AI trend in Asia the companies and the

interest on AI solution this part and

also my posive experience of the

Singapore. So I come here. Yeah. So in

the past two and two months I met a lot

of the customers partners and is a

fantastic team that is full about the

courage and the customers about the AI

and looking everyone is very is willing

and also have a very culture to to

embrace on the AI right now. So that's

really good here.

>> Yeah. So you mentioned a couple of

interesting things. One is you talked to

Tenfang about the development of LMS.

>> Yes.

>> Yeah. And you've I think you worked in a

couple of interesting companies back in

China that are already doing AI and all

of that. What are your views on how the

LLM space has evolved the last few

years? Yes. I think that from it should

be from the two years ago that we saw

that the LM is stuck from a moment from

the CHBT and that's a lot of company not

only startup company or big companies

they are training about the AI

technology involvement and in China

especially that in the two years you

have al see that they have accelerating

their pace about to how to adapt this

kind of technology into the applications

so I think that is very interesting and

I saw that a lot of company a lot of

industry company is trying to use the LM

to now is to changing their working flow

even change their B2 models

>> and then another thing interesting thing

you mentioned about talking to Tenfang

is like the difference between I guess

like AI development in China rest of

Asia

>> maybe you can share a little bit of what

you talked about with Tenfang right and

especially comparing that also to other

parts of the world what are the

differences in terms of how AI is

developing

>> yes of course that in the I think in the

China that right now they have the top

on that trend that all of the leaders is

trying to push about the new technology

into their business because they are

facing the more challenges about

business right about the growth and also

about the cost of saving but in Asia

right now company they normally the

industry person in customers they're

waiting for the new trading from the AI

technology data company and they need

the information from the like a Google

AI the platform company to get the

information together and they're waiting

for they have a lot of business case

they want to use AI but right now they

not exceecting AI solution fulfill for

here and also they will see a lot of

product is designed for China for

American but not here

>> and I guess that's what really attracted

you to whiz right because it's it's a

product designed specifically for

initially designed for Singapore

companies or Southeast Asia companies

right could you describe a little bit of

how your past experiences working in

tech in China have have shaped how

you're approaching your role now with

that AI

>> right now that I think we can talk about

the conversational AI what we is focused

on

>> to me that the conversational AI is not

only a piece about the technology but

also uh benchmark or even that itself is

a very core that core goal of the how

you say the artificial intelligence okay

because back to the story that you know

that the famous that the famous touring

test that's is a conversation scenario

right that we are supposing that the

computer or AI can persuade the the

people facing convert to him is facing

is it cannot be distinguished as a

human. Okay. What I see that okay the

conversational AI right now that to

development very very fast basing on the

large language model but we need to

notice that the conversational AI is not

right it's long before the large

language model just like basing on the

ASR TTS this kind of algorithm it have

already gain very steady and really

useful at the business team okay just

like the with solution we are already

have again a lot of the scenario use

cases in our industry so it they can be

used that in the not so open use cases

it's not so open conversation and it's

not required so humanlike convers

intelligence so this already gain the

business value right and it have scaled

into a industry so that's what I can see

okay

>> yeah I know quite interesting because I

think LLMs is just the latest

development in the well conversational

AI and you guys have been wiz has been

running since 2019 I guess long before

all these LLMs go into the mainstream

right

>> yes

>> what can you share about how wizai is

scaling out as a global company and how

is it developing a competitive mode

against especially today there are a lot

more AI companies

>> around and building I guess what how you

call like rapper companies right

building around existing LMS and all

yeah you're right there's quite a

challenge that we saw a lot of

challenges right of how competator is

going here and I think that we didn't

see that we see that is a positive

single because that you know that the

business values here okay and I think

the totally different is that we are not

right now is not startup from zero okay

with already a fundamental about the

solution and also the been the case so

right now when the large model

technologies here it help us to extend

to more use cases and to us that is the

marketing expand okay but I think to the

most of the competitor the new com

startup they are right now is working on

how to adapt a technology into the new

use cases so that's totally different so

we have a massive go ahead okay so

that's what we see in this what we are

seeing in this and right now we our

GTM's strategy is not focusing on the

top you know that we have a lot of the

customers is in big names in right here

in the region

I think they are quite innovation

innovation on the LM right now. So what

we are doing is trying to dig with them

about the use case how to get the most

benefit and most of the ROI and also

measurable value from B. So when we have

the mature about this case the solution

the methodology and also the change

management in this kind of tools. So we

go abroad, we go to the another

marketing, you know, go to a go globally

struggle right now.

>> Yeah.

>> So yeah, speaking of going globally,

like what are the learnings so far from

how different is it from China with

scale in Southeast Asia versus say

maybe? Yeah, I think that it's quite a

quite a lot of seeing the difference

between the different market right now

and I think that is it's a challenge

also that the positive point that we can

see that the here in Asia the different

country have different requirement and

also have the different governance about

the data about the hosting how to

hosting your server this kind of things

that's is a barrier or is a or it's the

most complicated power point to the big

companies yeah but actually it's for Our

startup like we we are agile and also we

are adapting and we grow up from the

regionally. So here we can get the

difference and we can know what is the

most efficiency application scenario and

what we can get the most benefit that

from the good data infrastructure right

now that we are building in this in

these countries. Yeah. So definitely

combine. And speaking of getting the

benefits from the different local

infrastructure, right? And your role is

really trying to build these stakeholder

relationships.

>> Yeah. How do you any learnings from that

point of view as a as somebody who's

trying to talk to I guess like

government, talk to other business

leaders and things like that? How are

you able to manage those relationships

and make them really benefit with that

AI? Well, I think right now that we have

a strong with a strong and also we have

the partnership that a lot of the

partner is helping up to talking with

these companies and also the government

that including that you know that in the

past few years we have did some project

for the government about very

complicated that cases that how to do

the online services for the customer for

the government. Okay. And also we did

that the transport like in central house

that's the nation host state there area

and we get a lot of like efficiency

effect here in a lot of pride here and

I'm interested to know what do you see

as future of conversational AI moving

forward LM is just the latest in a host

of different technologies that have

pushed conversational AI forward that

you mentioned TDS and all these other

things like what's the next step for

conversation AI

>> right now we are doing we are using the

lat model to benefit from our

fundamental product for example that it

will be covered that long tail about the

use case that is the requiring more

intelligence use cases yeah for example

that the inbound core that the customer

come in you don't know what is the

purpose right originally okay so it's a

very open scenario this the lockdown

will help us more

>> so our product right now is coverage a

lot of new use cases this the first step

and second step is that we found the

large

model will help us to stable all the

small language that the model from TTS

and ASR to have the training have the

material of the conversation material

this kind of things and to stronger our

process how to do the on boarding

process for us and second part we saw

that we the growth of the conversational

AI and also the how it both the large

model that a lot of capability will be

boost up for example that how you can

handle conversation with the customer

why he send you an image okay so this is

totally that use case we didn't mention

we didn't met before or we are not

required to handle that before but here

we are trying to emerge that all kind of

evaluate all of the technology growth

and also in embedded into our system

that is the road map what we can see we

can saw in the voice but all we can see

the pot right now.

>> So, it's opening up the different use

cases, making it more flexible, easier

to build new new scenarios or or even I

like what you mentioned about inbound

calls, right? The ability of the talkbot

to just assess already like what the

scenario is without it having to be

built for a specific scenario, right?

>> So many of the boundaries of

conversation content.

>> Yeah. So, instead of having like a

collection talkbot or customer service

talkbot, it's just one talkb that can do

>> Yeah. multiple kind of use cases

>> and practically that would have some

routter agent. Okay. That oh that can

transfer to another

>> right one agent and then another layer

of different agents. Yeah. Quite

exciting.

>> How do you see I think there's a lot of

talk right now about MCP's model context

protocols and how I guess it allows

developers to be able to off offload

costs right to to the large language

models and not have to spend so much to

to build internally. What are your views

on that data? It's a very interesting

question because the price model is

totally changeable and the s comparing

with the AI application with the

traditionally s software we have okay so

what I see and that they are charging by

the token cost or even by the how many

agent we replace about the head

>> I think I have some opinion that is just

a temporary to me that is temporary

because that what I see that if you are

calculating about the cost comparing

about the human that means that you are

we are assumpt we are assume the AI is

right now you're changing the workflow

right now the SE and working flow I

think it's not the future change because

that that is the fundamental change the

workflow will be changed the the

organization will be changed and even

the responsibility about the human will

be changed so I think if not a simple

compare about how you can replace the

headcom okay so I think that in future

maybe more that we're considering about

the bit of value in a new measure way

but not comp just comparing about the

cost to compare with crimate or the just

co model I don't think so

>> right so how do you think that will

affect pricing for a lot of these

companies and the business model for for

>> I think that the token cost away is is

very b is very good for the startup

company because allows to use the in AI

infrastructure you know very cost and in

in first step okay you know we can

invest a little to go into this area and

to make some application make the per so

that's good for right now.

>> Yeah. Allows you to run a lot more

experiments quickly and cheap cheap

cheap cheap as well.

>> How do you see wiz.ai in terms of I

remember two years two years ago

Jennifer was talking about her vision of

AI products was really flexibility right

and very user personalized solutions.

Where is Wiz.AI Yeah, in terms of that

heading towards that kind of

>> you saw that already changing a lot that

benefit from the lod model right now

chapter like you say that we can set up

that in different personality

>> and also in different language and also

the about the gender about to kind of

much more better about customer

experience right right now and what we

can see that maybe that we can make that

distinguish for each one just like you

see that water in the Yeah, a very

personal AI personalization. Let me

think this way.

>> I think that will be happen in the

consumer industry first and go in the

enterprise enterprise market.

>> Okay. So also will be happen here.

>> And what are the enterprise these days

looking for when it comes to AI

solutions? I guess many of them

especially the larger ones are already

familiar. Some of them are already ahead

like they're very innovative as you

mentioned like what are their biggest

considerations? I think the

consideration should be very familiar to

us because that it's exactly the same

with the what and what they facing in

the past year to facing a new technology

innovation. Okay. always including that

how you can measure about the ROI

>> and how can I get my business value to

unlock and how you can make the check

management defin not only a techn

knowledge change it's not a tool change

but you need to help us to activate

about what is the processing change what

is or organization change the event

model change so the consideration almost

the same but what we can see that maybe

something different in this time

>> because the AI is not different with the

previous technology. Uh the previous

technology always the push from top to

down. Okay. But right now that everyone

he use AI in the company that I think

from the intern to the CEO they already

using some non tools. So it's a it's a

really to be take us more confident to

make the on boarding process to the end

user. So that's something different

right now.

>> Yeah. Yeah. Does it? So what what are

the other remaining challenges when it

comes to selling to enterprise? What is

something that I guess stops them from

choosing one AI solution over another?

Are there still barriers to entry in

that case or

>> Oh, you mentioned about the buildup in

buyers. It remind me that I just read a

report from MIT that he called that the

AI and the general AI in business in

2025 and it named that sounds like the J

AI is divided.

>> So that is mentioned that they have the

finding the key finding is that 95% of

our organization is gain is getting the

zero return zero return from the from

the effort.

>> Okay. So they're trying to find the gap.

They started about the the builders

which more about the startup company and

providing AI product and they started

about they studied about the big company

that the buyer buyers okay and they

found that the builder is success

because their product is always focusing

on the narrow use cases and try to get

success in the just the use cases. Okay.

But the buyer, the big company, the

enterprise is always asking a more

complex solution. Right now they're

asking the product can be fed in their

specific workflow in integrated with

their existing tools and then

considering about their employees can

meet and the behaviors and also they

want to have the data boundaries.

>> Okay. So that's a big gap that the big

gap is just our air partner. So that's

what we want to build up to cooperate

with our our customers but we want to

make that bridge between this gap. Okay.

We know that they are not only need a

static AI product but also that your

long-term ship can help them to make the

change help them to adapt to the AI era

and help them to to go into change over

the time.

>> So that's what we want to provide in the

air.

>> And that's your role right?

>> Yes. Yes. We want to build off that

because that if you look back to see

what is we can provide that we are

already involved very deeply into the

customers working flow that build and we

also is build up on the AI technology

from the very early the conversational

AI and now they locked up with model we

know each other and also that we found

that we have and we are building right

now and we already have something we

have the tools to help to elevate

evaluate the business value we have the

methodology to help to make the change

and we have some product

>> already there so this is what we can do

>> yeah no quite interesting and I guess a

lot of the you have the wisd the AI

product itself but then you're also like

working with them especially the larger

like enterprise do you see these like AI

partners quote unquote also being having

agents for that as well instead of

instead of sales like a customer success

team like they would also work

internally with the platform.

>> Yes, of course because that what we see

on our customer that will be I think

that's two different kind of the drivers

of how to have AI. Why is that driven?

They are very focusing on how did the

use case how do you help me to do the

project to to deal with this use case

and measure that with the how to cut the

running cost time deployment cost the

times. In second part they're not only

to focus on one business case they want

to build an entire that AI architecture

for all of the department they want to

have the capability to build the agent

for different parts like I just

mentioned that the customer service

>> and internally study some reporting

analyze and some maybe some governance

and agent. So we help them to make that

the AI fundamental okay AI I call that

AI force build up okay and also that

evaluation framework dependency

>> and I guess before when it comes to SARS

I guess sometimes a lot of these

onboarding would take I guess several

months sometimes even a year like with

AI how does the turnaround usually is

how is it impacted

>> I think that it will be accelerated you

know but it's almost the same that also

went the same we did some P to prove

that what business value we can get and

then we you define about the solution

that what we need to change what the

system we need to integration with you

and then we go to the ending solution

the most major step is the same but

right now with the LM that the PC will

be much quicker and also that we can use

our appearance to make the agent to help

us make the methodology done

>> get a solution very quickly then for the

end to end solution also you know that

the the AI coding ran very strong out

for the interface integration.

>> So yeah.

>> Okay.

>> So would you do you know how much faster

it is now to to turn around? I cannot

measure about average depend depending

on capacity but what I saw that all of

our programmer already using the air

coding right

>> and they I think that maybe one to five

what you said seven capability right

yeah and as we the end of our chat like

I wanted to ask for your advice right

for leaders who are listening in who are

thinking about AI transformation as well

what advice do you have for them as they

if they're maybe experimenting with a

lot of different AI tools and worrying

about spend spending too much on all

these different AI solutions like how

should leaders think about experimenting

spending on these solutions and getting

the ROI that they want.

>> Yeah, I think that that's right now that

nobody is doubting about what is the AI.

>> Yeah, they can beat it right now. I

think the the right now the question is

that how you can adapt to AI as earlier

as possible. The first thing I think

that for the all of the leaders for the

industry enterprise they should thinking

about the AI architecture rather than

the technical architecture right now

here existing they should thinking about

that how to balance between different

mal for the enterprise perspective we

never want to bind into one vendor and

also I think that we can gain the AI

evaluations that just step by step from

the ED case from the most our Okay, we

can get to get the easy case. Okay. And

from the successful AI use case

employment, we can go to another one. So

just that that should be okay. And I

think that the most thing that I think

for all the leaders to thinking about is

that how the human is working with AI.

>> Yeah. Because that is the question for

even for us and we are thinking about

when the AI coming when AI is more

stronger and more the capabilities

increasing the use cases is increasing

that what is exact the working flow

right now what is the person the company

is need

>> okay so everything is changed so I think

for the leadership shift you think about

that kind of

>> and finally maybe you can share with our

audience what has been the most

challenging point in your career so far

that other leaders can learn from.

>> I think it's a big challenge that

they'll go overseas from the fast

growing China to see the the whole world

over the world because I think that the

the world is changing and the changing

is because of the the US and China the

this kind of

>> the this kind of complicity issues and

also the AI technology innovation. So

the two strong product is to make

everything changing. So I think that

should you you need to keep that tip in

mind to to see what is happening and to

adapt what you are doing right now.

>> Yeah. All right. Thank you so much for

thank you.

>> Thank you for coming on a call with us.

Yeah. All right.