João Diogo Falcão, Chief Technology Officer at AiFi

Julia Raymond Hare:
Today, I am super excited about the guest on the show. We are going to talk about one of my favorite topics, which was before the pandemic, autonomous technology and contactless technology, and even moreso now as we move into the rest of 2021.

JD:
Thank you. My name is João Diogo Falcão. Most people at AiFi call me JD, so you can call me JD.

Julia Raymond Hare:
JD. I’m so happy to have you on the show. So you are the Chief Technology Officer at AiFi, and your retail company helps power retail stores of all sizes through AI to become fully autonomous, which is super important today. You’ve advanced the company’s position in autonomous retail and actually created, as I looked up, the first AutoCheckout Competition in collaboration with Carnegie Mellon University and Stanford professors. Can you tell me a little bit about that before we hop in?

JD:
Sure. In my academic pursuits, I still maintain a close relationship with my old advisors, and I actually work with them and students to enhance the academic side of this challenge that is automated checkout. So I’ve created and chaired this competition where students from several universities came in and were given the infrastructure, which is something that’s hard from an academic to get into this field, and this common platform where they’re working together and they’re building their algorithms and looking at the data really to advance the state of the arts. This is one of its kind, because it really brings them all together into this one central platform that they can enhance. Otherwise, academics really find it hard to have access to data today, and access to the platforms, and to work and then compare each other and to really increase the state of the art. So that was the intention behind that.

Julia Raymond Hare:
That’s really cool. I love when I see tech companies partnering with the universities because it’s such a great way to get the younger generation involved and roll things out quickly that people might not have thought of. So before we go into a little bit about consumer trends and retail and the growth of checkout-free innovation, will you tell our listeners just a little bit about yourself and your journey to AiFi? How did you get to where you are today?

JD:
Yeah, sure. So I joined AiFi in 2017 as employee number six, I believe. I’ve always had a keen passion for technology from a very young age. I started coding when I was 10 years old because my brother was into that, and as I grew up in high school, I had a small business for repairing phones, and this really sparked the idea in me of becoming an entrepreneur. And then right around 2011, I started a company in Portugal that transformed the way facilities maintenance engineering was done. So in essence, we’ve built a system that centralized all of the building maintenance subsystems, such as elevators, fire alarms, et cetera, into a single smart, centralized platform. That was an amazing experience.

JD:
Since then, I sold the company. This was in 2014. After that, I decided to pursue a Ph.D. in the U.S. I did so at Carnegie Mellon University in cyber-physical systems. Now, this is just a fancy way of saying smart spaces or smart buildings or robotics for buildings. My thesis was in auto-checkout, which aligned quite well with what I was doing before in my previous company. During this program, I met very interesting people. One of those people was Steve, Steve Gu, our founder at AiFi, and it was love at first sight on his ideas, my ideas, and what he was doing. So very soon after, I joined AiFi.

Julia Raymond Hare:
Wow, that’s an amazing story. I can’t believe you started coding when you were only 10 and then had multiple companies before you landed the job with AiFi a few years ago. Because of your background and your Ph.D. that you got from Carnegie Mellon, how closely do you follow China, and how much do you think the trends in China will come across the way to the western world?

JD:
Well, I follow China actually closer for the reason that my peers and even advisor at Carnegie Mellon is from China, and a lot of the students are actually from China as well. So there’s a close collaboration there. So from an academic or technology perspective, I’m actually pretty aware of what’s coming from there, and it’s even more interesting that at AiFi, the co-founders are both from China as well, and we have teams in China. So there’s this close understanding from myself and from AiFi of what’s happening in China, and how can we leverage that what’s happening in the U.S. market, European market, and even there as well.

Julia Raymond Hare:
That’s really great to hear because I know that China is very far ahead when it comes to contactless technology. Some of them are cultural levers that are causing that, but are there consumer trends that are the key part of driving the growth of checkout-free innovation when it comes to the U.S., to Europe, to these parts of the world?

JD:
Most recently, COVID and the need for safe, contactless shopping has certainly influenced the speed at which consumer expectations are changing. However, the biggest trend that we saw even before COVID was convenience. E-commerce has led consumers to wanting products faster and closer to them, and there are two main points of friction for brick and mortar are location, so the distance that a shopper has to travel to get to the store, but also the checkout lines. So when you get to the stores, how long do you have to wait just to pay? Actually, Americans spend roughly 37 billion hours each year just waiting in lines.

Julia Raymond Hare:
That’s crazy.

JD:
Yeah.

Julia Raymond Hare:
Wow, Okay.

JD:
So at AiFi, we’re able to remove these points of friction by allowing retailers to open smaller, more efficient, unmanned stores closer to the consumers. Since we are a checkout-free solution, shoppers are never waiting in lines. So in the future, shopping for everyday groceries will be as convenient as walking to a nearby store, grabbing the product you want, and heading back home before Netflix auto-plays the next episode.

Julia Raymond Hare:
I will just bring up a little bit of the elephant in the room. Obviously, as someone who covers the retail industry from a media perspective, Amazon’s Go technology was huge when it launched in 2018, got a ton of buzz. A lot of critics were saying, “Well, this will take a generation to become economically viable. It’s super expensive.” It’s still expensive. What are your thoughts on this, on competing with one of the biggest players out there?

JD:
Well, I think that statement was certainly true in 2018, and that is because Amazon Go was the first of its kind. However, it’s our role, and that of startups in general, to challenge the status quo and essentially do more with less. This forces us to have to innovate and find these breakthroughs to survive. For example, just this past summer, we had an amazing breakthrough in our computer vision technology that allowed us to confidently move away from the status quo of using both cameras and weight sensors, or smart shelves, into a camera-only solution. As a result, retailers no longer need to install expensive sensors, and all the derived maintenance costs that come through that, which has significantly reduced the installation costs actually by 60% without compromising our accuracy.

Julia Raymond Hare:
That’s huge. That’s a huge breakthrough, a 60% reduction.

Julia Raymond Hare:
JD, why didn’t AiFi follow this camera only solution from the beginning or sooner?

JD:
Well, that’s quite an interesting question. When you’re pioneering a new technology or building a new system to solve a complex problem, the first step in this innovation process is to over instrument. What this means is to put more sensors in the store, add extra cameras, and be able to collect any other source of data that can help to understand the problem better. So all data points that you can collect from a real environment with actual shoppers will allow you to build a better system, and as such, the best approach was to over instrument to ensure that we were not missing any critical data pieces. If we had jumped a step, we would have lost the opportunity to gather this knowledge and data accumulation that allowed us really to get where we are today.

JD:
In parallel, and if you check our earlier pitches, we focused on simulation technology from a very early stage. This was done to gather and generate even more data. With all our investment in this simulation, it allows us to generate virtually infinite amount of synthetic data points, test a higher number of store configurations than it would be possible in real life. So our ability to combine this synthetic data with this real data that was gathered from the deployments with our partners and clients is really what allows us to deliver the best, most accurate, and affordable solution out there.

Julia Raymond Hare:
So it sounds like you said you combine the synthetic data with the real data, and create these combinations that otherwise wouldn’t be possible in order to optimize? And you said over instrument, is that the term you used?

JD:
Yes.

Julia Raymond Hare:
And that’s because were you using the data points from the, I guess I would say the old shelf technology with the cameras, to then fine-tune the camera only technology?

JD:
So indeed, having a store that has all these other data points, let’s say with the smart shelves as we were alluding to, really gives us things that camera alone doesn’t give us that needs to be trained and improve our AIs. I’ll give you an example. When you pick something from a shelf, if I asked you and you’re looking at the video, say, when did this person pick that object? You would say roughly around a certain window of time, right? They just picked it. A sensor can tell you to the millisecond, “This is the moment this person picked it.” So this type of information really helped us to then train our models, improve our vision models, to be much more accurate, and then you actually don’t need the sensors because now you know how to actually do it properly.

Julia Raymond Hare:
That’s amazing. I love that, that explanation. You did a great job, JD, of simplifying it down, how that process evolved to the technology you guys have today, proprietary technology. I’m just so surprised. The 60% reduction in cost seems like it’s becoming a lot more in reach for a lot of midsize and even some of the bigger retailers because we all know retailers have, they have tough margins sometimes. So this is a great way to push forward into the future with autonomous checkout.

Julia Raymond Hare:
Do you think a lot of retailers even realize this, that they don’t have to invest in smart shelves, that there’s camera technology that can do all of that?

JD:
Well, I think the marketplace has all these players, and each of them sending their own message that benefits them the most, of course. So it’s somewhat confusing today as a retailer, as a customer, to really know that there is such a solution that doesn’t use sensors and that it does work well, high accuracy, and affordable. So it is a challenge as a customer of this technology, and that’s our job to do better to inform them as well.

Julia Raymond Hare:
That makes sense. I will say, if I circle back to Amazon speaking with retailers, there is a lot of hesitation from licensing, their Just Walk Out technology. What are some drawbacks that you hear from your clients about licensing from Amazon versus going with AiFi?

JD:
Well, I can’t speak for all retailers, but one piece of feedback that was shared with me from our clients and partners is truly what you said, the concern of sharing data with Amazon. Now, Amazon is seen as a competitor for most retailers, and although I know Amazon has a strict policy on not using retailer data, it’s just an uncomfortable proposition for the retailers. However, I think that the main issue is the inability for retailers to customize their Just Walk Out technology and to fit into their desired customer experience.

JD:
We at AiFi don’t believe we know the shoppers better than the retailer. Actually, we believe that our clients can serve and know the shoppers, their clients better than we do. They can understand the value and their needs and desires and how this technology can better serve them. So we’ve built a platform, OASIS, you might’ve seen on our website, that is flexible to address the retailer needs to provide the best customer experience for the shoppers. That is, using for example their application or their discount cards or coupons, taking credit cards or not taking credit cards. So having unique store planograms or open doors versus restriction to members only, all of these options, all of this customization allows the retailer to shape the shopping experience to fit their brand. This level of flexibility isn’t actually built into Amazon’s offering.

Julia Raymond Hare:
I did see, when I looked at your client list in the past, Carrefour was on there and some other big names. That’s really impressive. Is there an example of a client that’s deployed your technology that you can talk about, for other retailers who are listening who might be interested in doing this?

JD:
Sure. Żabka is such a great example. They’re the largest C-store retailer in Poland, with over 6,000 stores and plans to deploy 1,000 more every year. It’s amazing. AiFi is contracted to install 35 stores in 2021, to further improve their shopping experience while driving the cost down. On those stores, shoppers can enter with their Żabka app, which has an amazing adoption of 3.5 million downloads in just two years since its debut, and then the shoppers can seamlessly shop with our computer vision technology, and when finished, just exit the store and get a receipt just moments after. Now, this is the shopper’s experience, but then there is this other side, which is the retailer analytics. We’re able to provide Żabka with a wealth of analytics, similar to what is expected from e-commerce in terms of path to purchase, basket insights, and even being able to tell how long a person held a certain product in their hand, or stood in front of it and ultimately didn’t buy it. So this really helps Żabka improve their product assortment, the way they planogram, the way they display their products, and even ultimately to adjust their pricing.

Julia Raymond Hare:
And these are things that AiFi can help analyze and provide insights on in terms of recommendations?

JD:
Absolutely.

Julia Raymond Hare:
Wow. If I’m a consumer listening, how would this work? Is it similar to Amazon Go where you have to have an account with that retailer and it’s through your iPhone or your mobile phone, whatever it is?

JD:
So this goes back to the earlier question, which is in this particular case, Żabka already has an application with amazing adoption. So we provide them the connection points, the platform. Now, as AiFi, if you’re a new retailer and you prefer to use credit card, or you prefer to use a different method of payment, our platform works with those as well. So it really depends on you as a retailer. How do you want to provide the best experience to your customers? The platform will provide that.

Julia Raymond Hare:
That’s great. When I think about contactless payment methods, because the vaccine for COVID has been rolling out for a while, and eventually people will go back to stores. But the sentiment that I’ve heard from most people on the show is there will be some habits that are long lasting that have developed from the pandemic. Part of that is just not touching things as much or not wanting to get too close to other people. So this helps solve that. Do you think that this will be something retailers can and should pick up right away? Because I mean, I still go into some stores, and I’m surprised to see they don’t even have mobile pay enabled. Right? And then in China, no one has used credit cards for years. So it’s a huge gap.

JD:
Yeah, absolutely. I think I can say more retailers should pick up this technology, and should pick up even digital payments. As an autonomous checkout technologist, I don’t choose between digital payment, whether it’s credit card or app, whether it’s in China versus Europe. Really, what we want to provide is this automation, this platform, so that the retailers can benefit from it, whatever the solution is.

Julia Raymond Hare:
Absolutely. When it comes to technology in general, and just as someone who is a leader in this space in retail, what are some things that you are most excited about as we move into the 2020s and beyond?

JD:
I believe that autonomous checkout can really transform the retail landscape in two fundamental ways. One is if you look at big box retailers, and you start to see this even today, they’re moving towards smaller stores. They’re moving towards placing these stores closer to the clients, in what they call the last mile to the client. This technology allows them to do so in a profitable way. The overhead of having a small store operated in a more densely populated area, which tends to have a higher real estate investment, is really compensated by the reduction of the labor required to run these stores. This is made possible through AI.

JD:
So I can see a demand for smaller format stores placed in residential communities, for example. Another fundamental way that autonomous checkout can transform the retail landscape is achieved through the need that brick and mortar stores now have to distinguish themselves from e-commerce. Ecommerce growth is forcing physical stores to fundamentally change the way they operate and shine through a more personalized and enhanced customer experience. So autonomous retail can allow the staff members that used to be in cashier stations to leave the cashier and focus on the store, whether it’s to make it more organized or cleaner or even spend more time with the customer to help them have a better experience. In the future, we’ll be able to connect online browsing history with real-time in-store offers for customers, but we can save that one for another time.

Julia Raymond Hare:
That’s a whole other discussion, right?

Julia Raymond Hare:
Thank you so much, João Diogo Falcão , the chief technology officer at AiFi, for coming on the show today. It was a pleasure speaking to you about a topic I’m really passionate about as I cover the retail industry. So super exciting stuff, and I can’t wait for the next time we have you on.

JD:
Thank you so much. It was my pleasure.