AI Everywhere

Introduction

Who I Am and What I Build

Hi, hello everybody.

Just to introduce myself, I'm Elisa Reina, I'm Spanish.

I am a veterinary surgeon and I got into AI like three years ago.

Basically how I got here is like I had a company called AyudaVets that was like help vets,

like helping vets to become better employees and better owners and that got me into automation

so we started doing AI automations for them and then we saw that automations were not

enough and we had to build something integrated, something vertical so we built our first tool

that was fully automated for them to manage their veterinary practices and then we have

been replicated in other sectors, therefore I had to build another company that is called

Solucionia because it was a bit difficult to go with a company that was called HelpBets to

other people that were not vets.

And also we have, I have another company called Laika which is more

AI advanced and it's an IoT device for pets that with sound and movement you can basically

basically register symptoms of the pets like vomiting, scratching, drinking too much, not

eating, things like that, and sends through the app a warning to the pet owner saying,

hey, go to the vet because your dog is getting sick or things like that.

So we have all that sort of range, so like veterinary sector, but also AI applications

into other verticals as well.

Also, I am a mother of two and I have to pick up a flight at nine, so at seven I will leave.

But it's very nice to meet you all.

Thank you for the opportunity.

Can you help me?

Okay, cool.

Why Niche AI Apps Are a Blue Ocean

So basically why I think it's now a blue ocean to build niche AI apps is because now everybody

can do it.

Look, I'm a pet.

Okay, I have an MBA, RDS, and my husband knows a lot about AI,

but I've been able to create it with a team of seven people that we are.

So with vibe coding, anybody can create an AI app for a sector in a vertical.

Now specialization is much better than scale,

so it's much better to build a customized product for, I don't know,

opticians or aesthetic doctors or whatever sector you know about

than a general one.

on.

And so now I think small markets are big opportunities.

Why Big Players Ignore Small Verticals

So basically, why do I think

there is like a blue ocean here?

Because like the big guys are not going to get into small

markets.

Markets are too small for them.

The high customization needs to be done and they

want a scale.

They want big products that they can implement in big customers.

They

They don't have the specific knowledge.

They cannot know about all the sectors.

So like, I don't know, the big four consulting companies, they know a bit about big companies

and big markets, but not about very much niche markets.

Also like there is regulatory complexity, especially if you go into health sectors,

which is my specialty.

So they don't want to get into the hassle of compliance with the different regulations

in the different markets because regulations here in Portugal are

completely different to the ones in Spain and if you go outside Europe it's

completely different as well so also for that even on the other hand we have the

good thing is to have like a small specialized companies because we have

the deep knowledge on a vertical we can have personal relationships we know that

market we we know the people in that market we know the sector I don't know

If I go to a veterinary conference, everybody knows me and they trust me and they're going to buy my product because they've known me for 25 years that I've been a vet.

The customization is agile because we are a small company.

It's not that when you have to change something, you need to, I don't know, discuss it with 20 people.

It's quick and that's what they want.

And it's also community building, creating industry networks.

What Small Niche Companies Need

works so basically what they want now i think in my opinion a small niche companies they want

a software that can help them with a company and it's compulsory that it has everything and ai

should be there and linked to ai because first i was building like ai phone calls ai whatsapp

ai bill automation and because it didn't get into their software it was not possible they were not

getting happy with this solution, 1so it has to be like a fully equipped solution.

Okay, so where I see the value proposition here, because you give them

a complete solution versus stitching multiple solutions.

If they have to go, I

don't know, for the clinical reports here, but then for marketing they have to go

to another software, for accounting to another software, then they get like, it's

too complex, everything should be together.

It has to be industry specific,

specific, affordable for small businesses, if they can get rid of the marketing company

that they hire, of the accountants, you know, even though if you charge a bigger amount

than normal softwares, it will be a big saving for them.

Competitive Advantages in Vertical SaaS

So what are the competitive advantages?

To have this deep vertical expertise, everybody knows about a market more than another, so

get into that market, get pre -built integration, so it's very, very important that this software

that you have because like people that has been building like sector softwares they started 20

or 25 years ago they don't have apis to connect to anything it's like a nightmare anything in ai

like phone calls whatsapp as i mentioned invoice anything that you want to help them with they

don't have apis and they won't allow you to input the data into the into the customers or the crm

and so you need to be AI power included

and a flexible approach

so be very open

and when they ask for something

introduce it into the software

so you make it live

and they see the value every day

because AI is changing so fast

you need to change with that

so with that you can change

that higher subscriptions

and then implementations fee

so you can have like a proper value proposition

Validating the Idea and Go-to-Market

So, why I think web app idea should work, but before, when you have an idea, I think

you should have a checklist to prove that this market will work.

Okay, it's like, okay, there is a market, yes, and is it underserved so people are happy

with their software?

Normally not, so there you have a market.

market?

Is it too small for big companies to get there?

Because, you know, they can

have like a huge ERP, I don't know, for Microsoft or something, but not something specific.

So is it too small for big companies to get into something customized?

Do they have customization

needs?

Do they have integration needs?

Because that is becoming very, very big nowadays.

nowadays, and will customers pay premium prices for something that has this access to AI directly

in their software?

So once you answer yes to all those questions, how have I done this?

I'm not talking about that this is the best method, this is how we have done it, and it

has worked so far, but I'm sure we will change it next month.

Stack and Tools We Use

So basically, we go for Replit, I guess Replit is like Globable, but in my opinion, like

a bit higher up into vibe coding.

So if you haven't tried it, try it because it's amazing.

I mean, you just do it in natural language and you create web apps in hours.

With that, you create like a nice front end and it makes you all the links to all your

your automizations with AI, so it's very, very quick,

is low -code maintenance.

Also, what we have done is link it to a big CRM,

an American CRM that allows you to be industry -adapted.

It's called GoHighLevel.

I'm sure most of you know it as well.

And then we create very vertical marketing tools

for that sector.

And then in accounting, we either use Holded in Spain because it's a very big Spanish company for accounting or QuickBooks if we use financial data for any other country.

Why does this help?

Because they have industry -specific accounting.

They have their KPIs.

You need to know these suppliers go to this part of the P &L.

so that helps them a lot to have everything together.

From Opportunity to MVP to Scale

So, okay, now we have identified the opportunity.

How do we start?

Okay, first is finding out that there is

a specific problem in the sector

that big companies ignore.

We have to validate it, so ask a few customers,

okay, how about, would you buy this product from me

if I build this product this way?

way.

Yes, once you have validated, okay, just build an MVP.

Seriously, if you go to Replit

or Lovable, Replit is more code -based because you don't need to add like a data set.

It

has its own data sets.

Just build the MVP.

Seriously, you can build it in hours and then

just test it.

Then you get your first three to five customers.

Maybe you give it for free

for a few months, then get all the feedback you can, and then basically just scale the

web app all the time and start getting more and more customers until you dominate the

niche.

So that's how we've done it.

Okay, so now we move into something else.

The Shift to Real-Time Speech-to-Speech AI

I know it's mixed into topics, but like last night we were trying to get the speech -to -speech

OpenAI from the real -time API from OpenAI engaged to our zip trunk to voice over IP.

So yesterday we managed to finish it, so I wanted to let you know about this because

I think this is going to be also a big change in everything, also in the way these apps

are built.

So, basically what this allows is to interact with computers and with all the programs speaking

and they reply speaking.

So before it was like speech needs to be translated to text, then text goes to the AI, replies

with text and then back to speech.

But now with this API, it's a speech to speech.

So basically, latency is zero and you can interact with machines in spoken language

and not in written language, which was code or like, you know, because it translated everything

into like natural writing, but this is really into natural language.

So this is a very, very big change and I'm very, very enthusiastic about this.

Why Voice Native Interfaces Matter

Why I think these things can help a lot because like it will give you like

immediate sentiment analysis so when you are doing a video call when you are

doing a presentation I could have it linked here saying come on guys guys are

getting bored or I don't know like they are not looking at you they are all the

time looking at the screen and then you know I would be able to adapt how I

communicate with you in real time.

I would get advice from this AI and adapt

how I speak to to you.

Also if I didn't speak English you know when I got

the invitation for this you said okay yeah you have to speak good English if

not you are not invited.

This is not gonna happen anymore because with this

speech -to -speech you can speak to it in any language and it will reply to you in

in the same language, immediately.

And in the same conversation, like I will love,

I will send you guys afterwards the number

so you can call and see it.

So it switches to another language.

So language is not a barrier anymore.

So anybody could be here in Lisbon talking to you

even if they didn't speak Portuguese or English or anything.

So you could have people from everybody, from everywhere.

Use Cases: Assistant, Coach, and External Brain

Also what I like a lot is that it's like an external brain.

okay this is also because I'm very related as I said to health and you know having somebody that

is advising you while you are doing a medical consultation is telling you hey you haven't asked

them what is it eating or does he sleep well well now you will have the baby you will see no it's

the interaction you you need to to to know a lot of data not forget everything so it's it's very

nice to have somebody in front of you before the customer so it it helps you

help the customer okay and also it's your personal coach during and after any

video call any interaction with any other people it will be your silent

analyst with very high impact I mean they will tell you okay you use I don't

know too much or too much this word or you could have been proven your your

communication this way or the way you were when moving was not correct you

know it will tell you like either in the spot or later on to improve yourself and

you can train to be a better you so basically it's a new era I think with

this speech -to -speech machine because it removes the friction between the

the language, the machines speak, and the way we speak.

So there is no friction anymore.

They can understand us directly speaking,

and they reply to us directly speaking without any latency.

So there is no more waiting for it to translate it into text

and then reply.

So that is fully over.

And this is a big, big change.

So it bridges the gap between text -based and real -time voice

experiences.

experiences.

The Future of AI Voice Interactions

So basically the future of AI voice interactions isn't just automated, it

is helping you to be a better you, more intelligent and more empathic.

That's it.

Conclusion

Thank you very

much.

Finished reading?