She Took On TikTok at IG, Was Tinder's CTO. Now this Founder is Betting on AI’s Next Big Opportunity

Name: She Took On TikTok at IG, Was Tinder's CTO. Now this Founder is Betting on AI’s Next Big Opportunity
Uploaded: 2026-09-04T05:21:22+00:00
Duration: 24 min 38 s
Channel: Entrepreneur

00:00He said, let Sean A answer all the phone. Basically, the phone didn't ring in the restaurant.

00:05Sean A answer like 350 calls. You're like, I never knew so many people called me.

00:18Welcome to Restaurant Influencers, presented by Entrepreneur. I'm your host, Sean Walchef.

00:22This is a Cali BBQ Media production in life, in the restaurant business, and in the new creator economy.

00:28We learn through lessons and stories. We travel all the way to the National Restaurant Show in Chicago.

00:3455,000 restaurant professionals, 2,200 vendors, 12 football fields.

00:39This is the biggest show on earth for restaurants.

00:42And our goal with this show is to find the best technology that is powering the next generation of restaurateurs.

00:51Maria Zhang, she is the CEO and founder of Polona AI. Maria, welcome to the show.

00:57Hello. Thank you for having me.

00:59It is so nice to have you. You have such an incredible background, former CTO of Tinder.

01:06You've worked at Google. You've worked at Meta.

01:09And now you are building AI tools for the restaurant industry.

01:15Let's start with CTO at Tinder.

01:18Okay.

01:18How early were you at Tinder?

01:20It was quite early. This is a company in West Hollywood, in LA.

01:25Yeah, when I joined, I was previously a VP of engineering at Yahoo.

01:30That was 2016.

01:33Wow.

01:34It was a small startup, but when I left, it really was quite a global phenomenon.

01:40How many users when you started?

01:42Yeah, it was even a million.

01:45A lot?

01:46Yeah, yeah, yeah.

01:47When you started, but what was the growth?

01:49Yeah, but on the growth, it was definitely hockey stick, right?

01:55It was not linear growth.

01:56And then particularly during my tenure here, we saw engagement improving, and also we started monetizing.

02:04Yeah, so now it's public information.

02:07I think Tinder is doing about $2 billion a year in revenue, all from subscriptions.

02:13So that was a part of enhancing that experience.

02:17We also built a lot of AI tools, maybe invisible to the end users, but content moderation, matching, recommendation, ranking,

02:29all of this.

02:30We actually won an award in 2017 for AI innovation back when I was at Tinder.

02:36That's amazing.

02:37Can you bring us to the Meta Days?

02:40Yeah.

02:40Tell me about leading the Instagram experience team.

02:44So when I joined, it was still Facebook.

02:46No Instagram?

02:47They hadn't purchased it?

02:48No, we have Instagram, but the company was still called Facebook.

02:52And then we went through the whole from Facebook to Meta, and I did work on Metaverse as well.

03:00When I was there, when the company was making that shift.

03:04And of course, now we're like, we're going to change the name back.

03:07Yeah, right?

03:07Yeah.

03:08But yes, I joined to do Instagram, and I was facing fierce competition with TikTok, and we launched Reels.

03:17We also did a bunch of AI work over there of content understanding, personalization, recommendations of the right content in

03:26the right moment.

03:26And then I moved from Instagram to a group called AI for Products.

03:32Okay.

03:32So they have AI for Infra, where my co-founder was on the Infra side, and also FAIR, AI Research,

03:39and my team was AI for Product.

03:41Yeah, so it was amazing.

03:44We actually just brought a couple leaders from the Gen AI team from Meta to join Polona.

03:50So definitely share some of the DNAs of moving fast, rig things, competitive, but also very customer-centric.

04:02You just glossed over the fact that Instagram went from a photo-sharing app to a video app with Instagram

04:08Reels.

04:09I mean, this is, you know, obviously TikTok came along, but every single person that's watching this, that's listening, you've

04:15been on Instagram Reels, and that's also Facebook Reels.

04:18How do you solve a problem like that?

04:20How do you, you know, work with a team to solve a problem to go from photo-sharing to video?

04:24It was quite intense, right?

04:26Because if you remember TikTok, which is one while, right?

04:31And so short-form video was still a new thing.

04:34Instagram's doing well with the feed and the stories, right?

04:39And then we actually build many different variations of the experience.

04:45I also think Reels is a great name.

04:48There were definitely other candidates of how do we call this thing, right?

04:52And how do you blend the original photo-sharing feed and then a very successful product, stories, and this new

05:01experience Reels together?

05:03And then at the time, we also have Instagram shops, the commercial, the advertisement, right?

05:08How do you blend all of that together?

05:11So it was a tremendous challenge, a very talented team.

05:15A team of hundreds of, I think, a couple thousand people working on that day and night.

05:23And it turned out to be, the product is getting better and better.

05:29It's incredibly successful.

05:29It's an incredibly successful product.

05:32But it wasn't a straight shot.

05:34It wasn't easy to get there.

05:36We definitely iterated a ton and made tough decisions along the way.

05:42And when I was there, I actually put together this little white paper.

05:46I still remember, I said, to win against TikTok, the secret sauce is AI.

05:52Why?

05:53Why?

05:54So, yeah.

05:55Oh, because it's all about, you know, understanding your customer, your user's interest, right?

06:02And also the temporal model of what's new, what's trending, what's here, what's now, and also understanding the creator profile

06:12and spot to the creator that has the potential of trending up and really understand, right?

06:18Every piece of content and every single user and create that real time.

06:24Yeah, so that's actually where we were already using very large AI models to solve these problems.

06:33Whether you're a global restaurant powerhouse or local independent like us here at Cali BBQ, we all want the same

06:39thing, sustainable growth.

06:41And let me tell you, PepsiCo is a partner built to do exactly that, help you build a sustainable business.

06:46As operators, we get it. We know that digital matters. Online ordering, third-party apps, digital menus, online visibility, it's

06:55a lot.

06:55And it's incredibly hard to prioritize when you're just trying to take care of your kitchen and manage a busy

07:00team.

07:01But a few small digital tweaks can have a massive impact, especially when you have the right team backing you

07:07up.

07:07That's where PepsiCo's Digital Lab stands apart.

07:10This is an in-house team built specifically for restaurant operators.

07:14They use proven media services and AI-backed menu optimization to drive what actually matters.

07:21More guests, smarter digital experiences, and more profitable online orders.

07:26Just look at the numbers from last year alone.

07:28And the absolute best part?

07:30These digital lab services are available completely free to Pepsi restaurant partners of any size.

07:36If you're a Pepsi customer, stop waiting.

07:38Talk to your Pepsi rep today or click the link in the show notes below to learn how to unlock

07:43your growth advantage.

07:44Let's keep that momentum going.

07:49What's interesting to me is that social media, when it started, it was who you followed and you wanted to

07:56see them in your feed.

07:57But you're talking about AI.

07:58And it's funny because as humans, we would think that, oh, I want to see the post that my uncle

08:03does and my cousin does and my friend from high school.

08:05But really, what we really want is what AI told us we want, which is what I actually watch, right?

08:13Yeah, it is.

08:14It's different.

08:14Like the human side of what we think versus the objective data.

08:18And the data tells us actually the experience is better once it shows me, Sean, you care about restaurants and

08:24you care about restaurant technology.

08:26Here are videos that you will like to watch.

08:28Yeah, those are tough decisions you have to make.

08:31Such transformation.

08:33You know, back then, before there was Internet, even in the early days of Internet, it was all editorial driven.

08:40You've got the magazine, you've got the newspaper, you even go on some of the websites, right?

08:44The editors in the back room, smoking scars or something, they were deciding this is the headline, this is what

08:50you see, this is everything, right?

08:51And that's now becoming, I think, of the past, right?

08:57Yeah, editorial content is still there, but I think the majority of content consumption is highly personalized and tailored and

09:06driven by the algorithm.

09:07But it is also a little scary, right?

09:10It's the algorithm feeding me and shaping me into the type of person I become.

09:18Can you share a little, like, why restaurants?

09:21Like, why did you start there?

09:22Why did you leave big tech?

09:23Yeah.

09:24I mean, we're talking about the biggest technology.

09:25Of course.

09:26You're building the biggest products that literally have global impact.

09:29Yeah.

09:29So I joined Google after my tenure at Meta.

09:35And at Google, I actually worked on, as part of Google Core, a set of tools to improve developer productivity,

09:45right?

09:45I think I always had a little bit of, like, how can we use technology to empower human beings, right?

09:52To empower professionals.

09:53And that was an amazing journey.

09:56And then now people talk about cloud code and so forth.

09:58And Google is still the very best at doing this.

10:01Right.

10:01But it just kind of down to me, okay, this is such amazing technology, agentic AI, generative AI.

10:09As a technologist, I see this wave of transformation is the most impactful.

10:16Even, you know, many, many times bigger than Internet.

10:21And then the iPhone came on mobile Internet, right?

10:25Or blockchain, what have you.

10:28I just feel a tremendous sense of excitement and responsibility, right?

10:35And how do I want to apply my skills, knowledge, expertise, experience to be most helpful and impactful?

10:43And, sure, we can help Google engineers be more productive, but I think they don't need a lot of help.

10:50Very, very successful company.

10:53And so the three of us came together.

10:55We're all computer scientists.

10:58And we couldn't think of a better group of people that we want to help more.

11:03And, of course, you know, I love food.

11:05I love cooking.

11:06I have friends.

11:07I'm in the industry.

11:08You guys are absolutely the hardest working people.

11:12We work hard.

11:13And there are many, many of you guys, right?

11:15It's the number one employer.

11:16It's a big, total addressable market.

11:18Exactly.

11:18And all these people, they're working so hard.

11:20And then we are also building AI models that are multi-modality.

11:25We're building AI that can see, that can hear, that can talk, right?

11:29So the coding code is just bits and bytes, right?

11:34And so, like, okay, right, let's build this very holistic, very agentic, like, full-modality AI systems and bring it,

11:45right, to the largest industry, right, in terms of GDP, in terms of a number of staff employed, right?

11:53The Food & Buff and Restaurant Food Services industry is the biggest industry.

11:57It's the industry I love.

11:59So it's kind of perfect all coming together.

12:03Did you know that Toast powers over 140,000 restaurants across the United States, Canada, and UK?

12:10It's an incredible company.

12:12I am on the Toast Customer Advisory Board.

12:14They are proud sponsors of this show, Restaurant Influencers.

12:17We couldn't do it without their support.

12:19They power our barbecue restaurants in San Diego.

12:22If you have questions about Toast, if you're thinking about bringing Toast on to be your primary technology partner at

12:28your restaurants, please reach out to me.

12:30I'm happy to get a local Toast representative to take care of you.

12:34You can reach me at Sean P. Welchef on Instagram.

12:38Once again, thank you to Toast for believing in the power of technology, the power of storytelling, the power of

12:43hospitality.

12:44Back to the show.

12:45So I think for me, the reason we do this show, we love our business.

12:50We love technology.

12:51We know that restaurants can run more profitable business, more sustainable business.

12:56When I met you and I met the team, we got really excited about the Voice AI product at the

13:01time.

13:02You trained our podcast.

13:03My podcast was Shaquille O'Neal, this Restaurant Influencer show, Digital Hospitality show, all of our podcasts to train Sean

13:10AI.

13:11And I was actually on the other side of the world with my wife in Istanbul on the busiest day

13:17for us as a barbecue business in San Diego, which is Father's Day.

13:20And we trusted you and your team to take our humans offline and allow Sean AI and Polona AI to

13:27answer the phone.

13:28Can you share that story with the audience?

13:30Yeah.

13:30First of all, thank you for your bravery.

13:32It was a lot of trust.

13:35And trust.

13:36But we also have a run test for a while, for a few weeks.

13:41And we had the confidence.

13:42And so we gradually turn it on, right?

13:45And that day you made the call.

13:47You said, let Sean AI answer all the phone.

13:49Basically, the phone didn't ring in the restaurant.

13:53And I think both of us were surprised that your Sean AI answered like 350 calls.

14:00You're like, I never knew so many people called me because the lines get busy and you never even picked

14:05up, right?

14:06And if you think about it, if you put a staff in the restaurant and the person would be doing

14:11nothing but answering calls, right?

14:12And of course, all of your staff were very busy.

14:15And not only every call is answered and they're asking you, is store open?

14:21Can I get a table?

14:22Do you still have ribs, briskets, right?

14:23And so more people come to a restaurant because every phone call was answered.

14:28And I think the takeout order doubled that day just naturally.

14:32And also the other, I think, pleasant surprise, right?

14:37It was like your table turned faster and the staff in the store can focus on the guests in front

14:46of them, right?

14:47They can sell more beverages or just take care of the guests at the table.

14:52So like all together, I did up to what, 18% year over year top line growth.

14:59Top line growth.

15:00It was absolutely amazing.

15:01We'll put a link.

15:02Adweek did a great job covering it.

15:03We'll put a link to that case study into the show notes.

15:07But let's talk about what you're building right now.

15:09Like what's getting you excited?

15:11Yeah.

15:11So when we built the Voice AI, we realized a lot of the calls were actually for large orders, right?

15:20Catering.

15:21And they're often off-premise and they need to order ahead.

15:25And it's an important event.

15:27It could be a wedding, reunion, conference.

15:29Yeah.

15:30And so I learned from restaurant operators, right?

15:34Such as a manager, a team.

15:36It's actually quite a complex and lengthy process.

15:40Yeah.

15:40So we actually built this AI catering manager, right?

15:44It is attached to your phone, but it's omnichannel.

15:47So we also can respond to your email, web inquiry, text, any form.

15:54We understand, oh, this is a catering order.

15:57We just launched our AI with the East Coast Mediterranean chain.

16:02And in one day, single day, $5,800.

16:06Come on.

16:07Wow.

16:07Single day.

16:08$5,000.

16:08$5,800.

16:09And that's pretty consistent.

16:11That's phenomenal.

16:11They do a lot of catering.

16:13And then it's just impossible for the human staff.

16:18You can have the best catering manager in the world.

16:21They keep really, really busy because they often also need to be in person to show up at the event.

16:27And then to coordinate, right, with everybody, the back of the house, the driver, and sometimes it also comes with

16:34services.

16:35Yeah.

16:35So you will never miss any inquiry, right?

16:40You always be the first because the AI responds instantaneously.

16:44And the AI will also remember to try, right?

16:48Here's proposal.

16:49We made some customizations.

16:51The AI is also consultative.

16:52Oh, what's your budget?

16:54How many kids?

16:54How many adults?

16:55What do you want?

16:56And they will learn all of the knowledge from the manager, from the catering manager, and then really guide the

17:04guest through the entire process.

17:06There is something else, right?

17:08So the AI also remember 24 hours before the big event to text you, email you any changes.

17:14And guess what?

17:16There's always changes.

17:16And you don't want to hear that the day off, right?

17:19It's just chaotic, right?

17:21So capture every changes.

17:23And then also follow up, right?

17:25After your event, probably if you're wedding, probably not 24 hours, but if it's a business event, 24 hours later,

17:33say, hey, how was the experience?

17:35Did you like everything?

17:36And then remind you to remind you to order again?

17:39Yes.

17:40Yeah.

17:40So it would generate repeat order.

17:42It's really just kind of this perfect catering manager.

17:45It would absolutely drive your catering revenue without you need to hire another staff.

17:51So I think what I come to this show for is for game-changing technology.

17:59Vision AI is something that you're working on, that we're implementing.

18:04I cannot wait to see the case studies that come out because I think it's just been so underserved.

18:11Security cameras in restaurants and what AI can do on top of it.

18:16Can you share a little bit about computer vision and how AI is going to make an impact in our

18:19industry?

18:20Yeah.

18:21So there are many cameras, particularly centered around the anti-theft, loss prevention, right?

18:31So it's been around for decades.

18:33Yeah.

18:33So what Polona is doing is quite fundamentally different.

18:38So we do have a, we call full modality model, meaning I can see, I can hear, I can understand.

18:45Under the hood, it's actually holistically one system.

18:48It's not a disconnected, separate system.

18:51One scenario, and I think we can implement it at your restaurant, right?

18:55Your restaurant gets really busy.

18:56And then when they call and they say, how busy are their table?

18:59Like, it's actually really hard to get that accurately.

19:02But now the camera can just take a look and say, oh, there is a high top by the bar.

19:07Just hop on in, right?

19:08Oh, there's like a few people waiting, but I think you can call.

19:11Well, I mean, it's interesting because Google has, like, these are the times that the restaurant's busy,

19:16but they don't have the actual seats in the restaurant.

19:19They don't.

19:20Which is amazing.

19:21Even with all the restaurants, it's almost never accurate.

19:24Or if there's a line, how long is the line?

19:25Exactly.

19:25Yeah, and often, like, maybe people just didn't show up.

19:28They cancel, whatever.

19:29So your reservation system could show that it's booked, but then the camera look at,

19:33no, there's still bars, high tops.

19:36Now there's one in the corner.

19:37You can sit outdoors.

19:39It really can give you this accurate information.

19:42That's just one example, but it can also just help you manage your restaurant the way you want to manage

19:49it,

19:49front of the house, back of the house.

19:51You define these events, right?

19:53Like, you tell the AI, go watch for if the trash can overflows.

19:57Go watch for if the staff is wearing gloves or whatever the problem you want.

20:03Barbecue sauce with the to-go orders.

20:04Exactly.

20:05Yeah, and then you can interact with the AI system, and then you can tell it, you did this wrong,

20:11or you can change events, right?

20:12So most of the other camera solutions are pre-defined, yeah, the theft, door open, et cetera.

20:21But for us, you really can just have this AI to help you monitor, to improve the guest experience,

20:30the operational efficiency, and also the performance of the staff, right?

20:35So you can coach them to do better.

20:38What do you think is the biggest misunderstanding with AI in restaurants?

20:45I think the biggest misunderstanding by my observation coming from tech is it's so many point solutions.

20:56Yeah, I have to say some very small solutions.

21:00I have a good friend, and she was teaching me.

21:03She was a very successful restaurateur, has like probably 15 locations in California.

21:09And I was like, why are you paying so many vendors?

21:12This one does this little thing, this one does this little thing, this one does it.

21:17And like I put all my software services on Amazon, AWS, right?

21:24Or I do not have 25 vendors, and then I hope to be that provider, right?

21:33We can give a holistic solution where everything is interconnected,

21:39where it can really offer an elevated layer of intelligence versus this one does tips,

21:49this one does update my, this one.

21:51It was tiny, and there were so many, I was like, how are you?

21:54And nothing was connecting, right?

21:56And it's just such a headache.

21:58You're already very busy running restaurants.

21:59Why are you having 20 little pieces of software doing little bitty of something, right?

22:04And they're like, this one's useful.

22:06I was like, cancel, cancel, cancel.

22:10Oh, that's amazing.

22:12Well, Polona is integrated with Toast, so that was amazing.

22:17The fact that you can call Sean AI, and you can actually order through Toast.

22:21I'm really excited about that.

22:22Maria, how can people learn more about Polona?

22:25Yeah, please do come to our website, polona.ai.

22:29We also have Instagram, TikTok, LinkedIn, YouTube.

22:35Yeah, we are trying to really share our insights and excitement and passion with everyone that's in the industry or

22:45has an interest.

22:46I am so grateful that you and your team, you picked restaurants.

22:49There's lots of other verticals that you guys could have gone in and solved solutions, but the fact that you

22:54started with restaurants, we are grateful.

22:56Please reach out to Maria and her team.

22:57They are absolutely amazing.

22:59We appreciate Toast for being the sponsor of this podcast.

23:02We appreciate Informa for giving us the opportunity to podcast from National Restaurant Association Show.

23:07Please come to the show.

23:08Please leave your restaurants.

23:10It took me 10 years to get out of my restaurant to come to the show, and this is my

23:13five years coming here.

23:14And I'm so grateful for all the people that I met, people like Maria, their team, so many incredible technology

23:19partners.

23:20F3 allowed us to be in their booth.

23:21If you need point-of-sale install or IT support, contact F3 Tech.

23:26Thank you for listening.

23:29Thank you for subscribing.

23:30We appreciate you, Maria.

23:31Thank you so much.

23:33As always, stay curious.

23:34Get involved.

23:35Don't be afraid to ask for help.

23:36If you need to reach out to me, I'm weirdly available at Sean P. Welchup.

23:39We'll catch you next episode.

23:45Thank you for listening.

23:46If you've made it this long, you are part of the community.

23:49You're part of the tribe.

23:49We can't do this alone.

23:51We started.

23:51No one was listening.

23:52Now we have a community of digital hospitality leaders all over the globe.

23:57Please check out our new series called Restaurant Technology Substack.

24:01It's a Substack newsletter.

24:02It's free.

24:03It's some of our deep work on the best technology for restaurants.

24:06Also, go to YouTube and subscribe to Cali BBQ Media.

24:10Cali BBQ Media on YouTube.

24:12We've been putting out a lot of new original content.

24:14Hopefully, you guys like that content.

24:16If you want to work with us, go to etheshow.media.

24:19We show up all over the United States, some international countries.

24:23We would love to work with you and your growing brand on digital storytelling.

24:27You can reach out to me anytime at Sean P. Welchup on Instagram.

24:30I'm weirdly available.

24:32Stay curious.

24:33Get involved.

24:33Don't be afraid to ask for help.

24:35We'll catch you next episode.

She Took On TikTok at IG, Was Tinder's CTO. Now this Founder is Betting on AI’s Next Big Opportunity

Category

Transcript

Recommended