Fatima Awan
Obviously these big giant companies, they have the resources to be able to do that. But how does anybody out there? Or how does even entrepreneurs with budding ideas that are going into the business world, how are they able to tap into these technologies, and I think that's where companies like us come into play so that they can also take advantage of it.
Finola Howard
Founder of finite AI, an AI solution for the skincare industry, we're used to hearing about language models for AI now, but in this case, Fatma is using computer vision, which is a type of AI as a way to interpret skin types, helping consumers buy the right products for their skin type, and skincare companies, deepen their relationships with their customers, and increase sales at the same time. So let's get stuck in and find out more. I'm Finola Howard, intuitive marketer, your host and founder of how great marketing works, I believe that every business has a story to tell, because that's how the market decides whether to buy or not. And your story has to resonate with who you are, and with the people you want to serve. And this podcast is about helping you reach the market in a way that feels right to you. So if you're an entrepreneur with a dream you want to make real, then this is the podcast for you. Because great marketing is your truth share. And today, I want to introduce you to someone very interesting, who's doing some unique work in the space in the AI space. And I wanted to share with you a different approach to entrepreneurship, or the startup tech approach to entrepreneurship that we get someone really early on in the process, who's already had some successes in a really new space. So today, I wanted to introduce you to Fatima r1 from finite AI, and we're going to find out about her story. Welcome Fatima.
Unknown Speaker
Thank you so much for having me.
Finola Howard
I want to share with everyone how we actually connected which is quite bizarre in some ways. And maybe it's not bizarre at all. Something happened we on Twitter, we actually connected through Twitter.
Speaker 1
Yeah, that's right. Through our main brand account, which is finite AI, you know, we were just kind of trying to engage with our end audience, which is basically skincare brands or E commerce and retailers. And we came across an EM from Harlow, the US. Yeah, yeah. And we saw that there was an engagement. It was her podcast that was actually with you a while back. And yeah, we saw that, you know, we learned about her story. We're like, well, this is this is who we want to be speaking with. And so she was very interested in in that connected that way. And then I
Finola Howard
got really interested in. I wanted you on the show. So it's really fantastic. Okay, let us start with with start with your story. Yeah, let's start with where maybe tell people about the business first really quickly, and we can get back into it. But then tell us about how you started in your background? Sure.
Speaker 1
Yeah. So with finite, it's basically, we are an AI software. We're a tech startup, we started back in 2020. I mean, can get into the journey of it. But basically the product we have today is an API solution for online retailers and E commerce platforms to help them increase their sales conversion rate. And we do that through our deep learning personalization software. So basically provides recommendations to customers as they are shopping on any given site.
Finola Howard
Yes, but there's more than that. That really, you're so used to telling the story. So I wanted to say a little bit more. Okay. So what really amazed me was we all think of AI from the perspective of Chachi PT, and language models and things like that, but what really interested me was you had figured out a way that you take an image and see and look at the pixels of somebody's skin and based on what on all of the data that you have accumulated over this last while based on that data, be able to recommend skincare products based on the accumulation of pixels in their skin. Is that correct? It
Speaker 1
Yeah, kind of. So real layman's. So basically if we talk about the technology itself, we are looking at computer vision. So you know you mentioned about generative AI and we have seen last year in January there was an explosion of AI basically because if open AI is chalked up to release, and mainly you will notice that even within e commerce and marketing there's a heavy focus on generative AI and you know GBT based technologies, what we are looking at is slightly different, we're looking at computer vision. And, you know, even deeper within that is CNN. So that's convolution neural networks, there is a slight difference between the two, and I'll get into it. But with computer vision, what we're looking at is, you know, detecting information from an image. And when we talk about an image, it's basically a grid of pixels. So the way that we have developed this technology is we have trained, you know, a data set of about 40,011 100 original images. So, you know, looking at millions of pixel level annotations based on the RGB values, which is, you know, red, green, blue. And so, you know, you can basically train and build an algorithm as such, so that the difference between, you know, the magnitude, or the difference between the skin textures is in the difference between the magnitude of the pixels based on their RGB values. Now, with Gen AI and CNN, the difference is, I mean, they're both machine learning, you know, methods, but with generative AI uses something called neural net transformers, like, you know, generative adversarial networks. So, they will essentially generate or discriminate a task, and that's something you will see chat GPT doing, you know, it will generate tax based on, you know, an input that you provided, or generate an image based on an input text input that you provide it. Whereas CNN is basically does image recognition and classification. So, you know, one will basically recognise an object, the other generates it. Okay.
Finola Howard
Thank you for clarifying with your use of image and the code, as you put it, computer vision, tell us specifically about that. So I get the sense that if it's looking at the RGB values of skin tone, texture, etc. So say, for example, I have rosacea, which means I have red skin red, it says in my skin, that then I'm presuming means that you can look at a set of products that will actually recommend skincare products that suit the realisation in my skin.
Speaker 1
That's right. So, I mean, when we're talking about the RGB values, that's getting very deep into the actual process of how the algorithm is trained. But fundamentally, if you look at like, the concept behind it is, you take a set, you know, image set, you tell that image, like we did classification, and segmentation with classification is just a label on the image that will say, hey, this image has Horizont it, for example, or acne on it. So it's like a text label on the image. Now, that's great. But you know, within that image, there's, you have a whole, you know, facial structure. And within that there's, there can be a lot of different other factors going on as well, you know, in parallel to maybe pores, or even within the image itself, where do you identify the regions of pores, so we have actually gone a step further to that and segmented it. So we created these maps within the face where, you know, it's looking at areas of pores or areas of acne, or, you know, dry areas or wrinkles, or so on, and so forth. So, once, once you have that labelled and segmented data set, then you can build the algorithm and then it will basically provide an output based on the learn information it already has. So the output that the AI now gives, so the way it works is, you know, you go on a shoppers website, you see a lots of products, like, for example, let's look at Amazon, and you want to buy a sunscreen, you know, you can type in sunscreen into the Amazon search bar, and you're probably gonna get some, like 500 pages of sunscreen pop up, because Amazon's a giant platform, there's so many products out there. Now, how do you know which sunscreen is cracked for you, you can filter it, there's filtering methods on the side of the of the of the platform itself. And even like with most ecommerce stores, as well, you know, you can filter based on tax, or you can filter based on the options they have on the side. And that's going with the presumption that you know what it is that you're looking for. But if you don't, then you're kind of like just browsing through many pages of products, and maybe just, you know, making a guess, and then buying the product and then using it but maybe not liking it. So we've tried to like, reduce all of this friction within the shopping experience, where now the consumer can just upload a photo or take a live photo, and the AI does the detection for them. And then it will show all of these recommendations that are suitable based on what the AI has detected within the image. The information that it will give you on the outset is tech space. So it'll tell you so it basically is digesting information but giving you something that you can actually easily understand. So it'll tell you hey your skin is looking Normal or Hey, your skin is looking dry. And we have detected areas. Of course, these are some products that you know, can help address that. So that tastic
Finola Howard
Yes, excellent. Because we have to make it so people can understand it. And that's just how you've explained it feels so simple. But it's also quite mind blowing. You know what I mean? It's really Bravo. To that the thing that I love about your story, Fatima is, this was a problem you had? Yeah, that's right. You became your own guinea pig for your own tech startup?
Speaker 1
Yes. All right. Yeah, that's very right. Why
Finola Howard
don't you tell us that story, because the other part of the story that I like is that you've, you've looked at two different business models. So tell us the story of how how of the challenge that you had, personally, and how that turned into a business. And then the two approaches you've had in turning that into, you know, two different revenue streams?
Speaker 1
Sure. And so that would take me back to when I first started the business, which was back in 2020. And, you know, if we go even further back to that, you know, I was working in New York for another digital health startup, I was running their r&d. And that's kind of ties into my background as an engineer. So I knew it was kind of setting me up for a stage where I wanted to pursue a business of my own, you know, figure all of those things out, it was just teaching at me, but I wasn't sure how I was going to do that, what it would entail, I was quite freaked out, I didn't even know what I was going to do. But, you know, mid 2019, when I moved back from New York, back home in Canada, you know, kind of toying with the idea of starting my own business, not sure where to start, I came across founder Institute, which is a precede accelerator programme out of Silicon Valley, but they have offices all over the globe. So they had a location in Waterloo, and I think, I believe that was relatively new. So I joined the programme, you know, we started off with, I forget the exact number, but there was a lot of us, maybe 100 or so of us, you know, budding entrepreneurs with ideas, wanting to refine the message of it. And at the end of it, we only had about 11 companies that actually formed after the four months. So we were one of them. Basically, it allowed us to really kind of, you know, settle in on the one probe the one idea and kind of test it from all directions to make sure that it works, before we go out and launch this full fledged company. And at the time, you know, when I'm thinking about what the business is going to be, I'm also struggling with and thinking it was actually a ha moment, like, oh, my gosh, why isn't this happened, because, you know, in the background, I have kind of struggled with like skin issues. And mainly because, you know, I would go to the store, and back then there's not that much online, but there is a lot of brick and mortar. So you go into the store, you see 1000s of products on the shelf, you make a purchase, you come home, you try it on, and then I would have like adverse reactions to it. So it was starting to get annoying where it you know, it's like you're, you're spending money, you know, you're putting in the energy to buy the product as a consumer as you should have the way you have been advertised to, but it's not working for you. So I wanted to be able to automate that process in a way that could work for that would simplify it for the end consumer. And obviously, with our phones, we already have, even back in 2020, facial recognition, and even prior years before that facial recognition already existed. So it wasn't something new or novel. But we just kind of took it a step further. So we took a basic commodity like our phone and turn it into a way where you know, you can take a photo and the photos are matching with product recommendations. So when when I first launched in 2020, early 2020, the idea was an app, a direct to consumer app, because that's how I had pictured myself using it. So we did we got we went ahead and built the app, it was Bootstrap. I use my you know, my own funding. And we launched the app to the Apple App Store and Google Play. It even exists there. We have some 10,000 user downloads. And you know, the app was basically finding products from Amazon, actually. So even now, you know, the users, they take a photo and then it will match you with products on Amazon. So you know, the fact that you know, new users are actively downloading it without spending any marketing on it is is in itself a validation for me. So I'm happy about that.
Finola Howard
antastic and it's called the finite app, or what do you what did you call it? Yes, it's
Speaker 1
just just finite. You can find it on on the App Store or Google Play and give it a try. Yeah, but with any kind of business, you know, evolves over time, and we saw that there was an there was a need, much larger than what we had originally anticipated. So in 2022, we raised our first round of funding. We received a precede note from Murrell accelerator, which is an investment firm out of San Francisco Silicon Valley, and had a bit of a pivot. So now the focus was b2b, where we could develop a product that could be applied to the larger and broader market with retailers take a moment, because
Finola Howard
I think it's really interesting. What made you realise you needed to do that pivot?
Speaker 1
I would say it's a combination of a bunch of things, but mainly what I was learning through that journey of building that initial product, which was the app. Yes, we built the app, and it was for the consumers. But I was seeing that this technology could be largely applied to the broader markets. And also when I was, you know, interacting with different stakeholders, like investors or, you know, just pitching to people in general, the the feedback that I was getting was, well, what about the brands? And what about the E commerce? And what about the retailers because they also have a need for that, and a really big gap right now in the market for that. So it was always playing in the back of my head. But at that time, which was the early years of this idea, I was first focused on, you know, even getting the technology right, and building this initial product that we had envisioned, and especially with the idea of, you know, doing it with Amazon, it was just also appealing. So very much focused on that. But yes, it was it was playing, it was brewing in the back of my mind. And then finally, it was the right, I think, time in 2022, when we got the funding, and now we could heavily focused on the b2b side of it. And, and yeah, go from there.
Finola Howard
So what does that look like? So your your b2b, tell us about the b2b? So how, how is that going to work?
Speaker 1
So with the b2b you we had, when we were thinking about the product, it had to be simple enough that it could be applied. I mean, ecommerce is a huge market, if we were just talking about skincare, particularly within e commerce, it's $183 billion. And it's projected to be even more than in the next decade. So you know, when we think about this market, we had to build a product that can be easily applied to the end customer, without there being too much friction in it. So what we came up with was an API solution. So with the API, what I mean is, we have five endpoints, these are just like URLs, which you can embed within any ecommerce platform like Shopify WooCommerce, or if you're using your own technologies, like Magento, Wordpress, whatever. And you can enable and tap into our machine learning models and, you know, enable that personalization for your end customer. So with this API, what it does, basically any kind of the value proposition for the end customer is that, you know, we built it in a way that it is modular, and as I mentioned, platform agnostic, but what we're focusing on is for the end customer to be able to increase their sales conversion rate, because now that the shopper has kind of, you know, really tapped into their needs and are making purchases based off of it, it is, you know, thereby going to increase your conversion, it is going to increase your customer engagement ad is going to have that repeat loyalty value to it. And we went a bit further than that, because once we launched, once we built this product, and launched it last year, in 2023, you know, and got those first customers on board, we were starting to, you know, take in the feedback from them, and like, you know, the product that we had delivered to them what it is that they were kind of the missing aspects of it. So that really, I think, was great to what the product is now, because what we were learning is that, beyond the personalization and beyond, you know, that initial ROI for the customer, which is the conversion and the engagement and the loyalty, they also wanted, you know, as a skincare company, or as a retailer, they wanted to get data enriched company. So, we were, I mean the data is being generated, but they were not able to collect it. So we built a way for them to now be able to collect that data on their end as well so that they can now have or tap into more informed decisions. So now, basically, what happens is, if you know someone is going through that process on on a retailer's platform, that data gets sent on the back end, they can tap into it on the dashboard, and they can download it and and connect it to their CRM so they can actively, you know, get involved in marketing campaigns based on data that is being generated on their platform, which I think was a huge value add for, for our customer.
Finola Howard
huge value add. Yeah, amazing. So just so that we can explain to people so now, in this you pivot that you've changed it to or you've added to the b2b perspective. Now, what you've got is a product for a skincare companies piece of code that five pieces of code that they can add to their website, that when a customer comes to their website, that bad customer of theirs will upload a photograph of theirs. And, and what will spit back to them is recommendation of their products, just their products for that customer. And as a result, presumably from a marketing perspective, you tighten that relationship with the customer, because you have that personalised approach because you've recommended something for their skincare, which automatically brings them closer as a customer. Yeah, that's fantastic.
Unknown Speaker
That's correct. Yeah.
Finola Howard
So good. I'm glad I got that. Right. So and then the other part, which I think is really interesting that data enrichment that's because clearly, that's something that's we're all hearing about day to day to date, and we've heard about it for years. But now with AI, it seems to be that it becomes more democratised. We're all actually being able to access data for our businesses, our data, because that, who will there's my question for you who owns the data? So
Speaker 1
while it depends on what you define as the data, are you like talking about the images? Or are you talking about the feedback from the AI? I guess the short answer to that is, it depends on the client, whether because we see both sides of it there, we get clients who don't want any part of the data because of, you know, especially for talking about European clients of GDPR reasons. So they're like, We want to be compliant with those laws, we don't want any of the data state. So if that's the case, then none of the data is saved. And none of the doesn't go to that route. But if if the client is looking to save the data, then or they choosing which parts of the data they want to save, then they can do so like, whether that's the image if that's what they want, or if you know, the AI feedback that's being given from the from the consumer and from the images. And also, with the CRM integration, what happens is, you know, now when the user uploads the photo and sees the product recommendation, they can also, you know, give out their email so that they can be contacted later on. So all of that gets synchronised into one dashboard. So you have the email with the customer, you have the customers information, which is there's a time scale concerns. So now you're not only just, you know, showing products to your customers, but you're learning what it is that the customers want, or what their needs are, and it helps you kind of, you know, now, in for future r&d, like you want to, you know, focus on areas that your customers are interested in, or your customers have a need of sight provides demographic information,
Finola Howard
I presume, then you'll also start to see patterns of the types of skin that are most buying your products where you're most successful with. And it's actually not just salutely sales figure, it's got, you've got a richness to the data. So it's not just, I sold 10 boxes of this, it's like, no, I sent 10 boxes of this. And these are the types of people who are doing it and bump a bomb. And it's not what we expected, you know, can be things like that will happen.
Speaker 1
That's absolutely right. And that's exactly what's what's happening now, as a result of it. And you're seeing, and it's interesting, because you mentioned about democracy and democratic, democratically, you know, building products that can be applied to everybody. Because the reason I say that is, if we look at large platforms, like Shopify or Amazon, these are I mean, they have a huge market share. And all these leads, they understand the need of AI. We're not even talking about the future anymore, it's here, it's in the present, they understand the need of it to solve their business problems and to be still relevant in the market. So they have already developed their own tools. Like for example, Amazon developed a programme called Sequoia, I think we're now they're able to not just it's not just the Gen AI that's writing the product descriptions or, you know, enabling better Smart Filters, or even product recommendations based on text. It's now able to, you know, tap into and look at inventory data. So they're able to process and reduce processing times, or they're able to plot delivery routes, you know, based on real time traffic or weather data, or they're able to predict the demand for products. So they have 400 million products, but now they can do a daily demand prediction on what kind of products customers want to purchase. And not just like in one particular area. It's like all over the world on a daily basis. I think that's huge. But obviously these big giant companies, they have the resources to be able to do that but you You know, how does? How does anybody out there? Or how does you know, small, even entrepreneurs that are like, with budding ideas that are going on into the business world? How are they able to tap into these technologies? And I think that's where companies like us come into play so that they can also take advantage of it.
Finola Howard
And you realise that's what you were doing when you started? No,
Speaker 1
I did not. know, honestly, I didn't I when I started, I, it was just okay. Yeah, an app, let's automate this. And I was I was, I have to be honest, I was very focused on the mobile part of it. Because we all use our phones, so. And even now, with the b2b shift, you can use it on your smartphone. So like with all of these skincare brands, if you're using our product, your customer can tap into this on the mobile to also. But yeah, when I started off, it was just an app. And it was just an app that was linked to Amazon. Now, it's not just that it can be linked to any e commerce platform.
Finola Howard
Let me ask your thoughts now about you know, there's there's two, there's two divides going on with AI. And one is, oh, my God, look, we can make things so much easier, and so much simpler and do things much faster and really progress. And then there's the flip side of, we're really scared about AI. What's What are your feelings about that working in this space?
Speaker 1
In the working space or in general?
Finola Howard
Know what I mean, you're in this space. So your business is called Five Eyes. So you will have
Speaker 1
to, I guess, the way I approach it is how I approach anything in life, I guess I am a centrist. So I like to have a balance. With AI, yes, there's a lot of talk around Oh, yeah, you know, with AI is going to basically replace humans in the workplace, and it's going to replace everything, everything's going to be automated. There is some truth to it. But the technology can also be supplemental to what we're doing. So and that's the way I'd like to approach it. And I think, you know, anything, like if we were back in the stone age's, and we were thinking about a computer that could like now, you know, like with Excel, or any of these Microsoft products that came out and make our business world so much more easier. Before that, we were like manually doing tests, but with these products, it made it more efficient. So that's the way I see AI. I don't think it needs to be too crazy. But I think if if it can supplement your work and make it more efficient, and make businesses stronger and more competitive in the marketplace, then then that's a good thing. But is it gonna replace us? And are we going to all become cyborgs? And machine bots and whatnot? Hope not. But who knows? It could also happen, I think the way I would say is that have a have a balanced approach to it and figure out why it is that you're tapping into the AI technology. You know, what are your end goal for it? What are you looking to get out of it? And so I think it depends on the individual? I think it depends on the business. What it is that they're looking to do with the technology?
Finola Howard
And how are you? How are your b2b customers now? Are they getting their heads around it? Are they excited by it? What were they?
Speaker 1
Especially with, as I mentioned, with GPT, once it came out, everybody was like there was an AI uprising. So it's it, it's no longer like teaching the customer that the pain points anymore, the customer has already is aware of it. So it's much more simpler than that. So yeah, I think it's easier, especially with open eyes release. Got, it got a bit more easier than that.
Finola Howard
In some ways, I think what it's doing is, as I reflected, the customer is actually it's not we're not asking about pain points anymore, or coming up with ideas for solutions for for our customers, the customers are telling us and we're just simply acting on what the customer wants. And it's the most direct route now. Yeah,
Speaker 1
that's right. I mean, so like with our customers, you know, you have these brands that are already doing the digital marketing, they're spending money on, you know, Facebook or Instagram ads or even in content marketing. So if, and they might be using generative AI for that for content marketing. So there's a lot of play on that already. So they understand AI from the generative perspective, which is what we're talking about earlier, the difference between generative and CNN and computer vision. So we're just an additional layer on top of it. Where it's, it's, it's not based on just text and it's not just based on like advertisement. It's the AI is basically doing all of that for you. So like once it's implemented on the website, it auto markets the product to the customer, and it also takes the customer's information so that you have better informed decisions in the future or even in the present. So it's basically a smarter way to go about how you would market your brand, and how you position your brand to your customers.
Finola Howard
Absolutely. Tell me share with us, if you don't mind, high point in a low point, as a startup, our biggest lesson,
Speaker 1
you know what, I don't have any work. So we don't, I wish I could sit here and be like, oh wise about it, it's not easy it is. It's not as glamorous as it's made out, you're at least when I was, you know, first starting out, you know, you've heard you hear all these stories about the early 2000s. And the big rush of the internet, and all these like Silicon Valley companies starting now. And you hear success stories of these founders, you know, PayPal, and Facebook and all of these big companies started about, you hear about the venture capital world and all of that. So it's cool, but it's evolved from all of that, to what it is now. Now, anyone can start a company, you don't need venture capital, you don't need funding, you know, even because, with all these new technologies, you can literally build a piece of code and, you know, have the product ready and go to market with it. But But entrepreneurship in itself, I think it has, it has become easier because of the technologies. But I think fundamentally, it is still it's a very lonely process that I will tell you, it's very lonely, you're, you know, you have to really believe in your vision and what you were kind of kind of imagining for your business to be able to do it in the long run. So, yeah, it's tough. But I think if you stick around to it, and to stick around to it, let me add, let me add to that as well, to be able to stick around in a journey like that, I think you have to have some sort of level of insanity, because a rational mind would not keep pushing you forward. So if you have those things, go for it.
Finola Howard
But I think, you know, lots of entrepreneurs say that bit of insanity, or also they say ego. But I also do think that entrepreneurs have purpose. And that's okay. Yeah, knowing that they want to solve a problem. I do. I just see it too much in every conversation I have. And always they're self effacing and saying, oh, no, it's my ego or being I'm crazy. But no, it's, it's a drive to solve a problem.
Speaker 1
That's right. And I think yeah, I mean, if that's not there, then you I mean, you're just not gonna make it past the initial like, one week or something, it's just not going to work. I think for me, what kept pushing me was, I mean, inherently, I'm a builder. I'm an engineer. So I love to build. And also, you know, on the side have gotten into construction of it, too. I love noticing, there's other parts of me that I that I like to apply that too. So I like building. So I was excited to build and solve this problem. And that's what I set out to do. And then it evolved, and it became another thing. So then I started set up to go build that. So I think one after another, it was kind of an iterative process where I was building on, you know, on pain points on on things that needed to be addressed within the market, where there was a gap. So yeah, so I think that that's what kept me going. But definitely, it is not an easy journey. But the one that I am glad I went on.
Finola Howard
That's a good way to finish. One that you're glad you and done. And what is any thing exciting coming for next steps for finite
Speaker 1
AI? Well, we are, we have been through the pain of building the products and now we're just happy to you know, deliver it to our customers. So we're focused on string our relationships with the customer solving, and addressing their pain points with our services. So yeah, so you know, growing the business. And then after this venture, we'll see where we're headed beyond that. Because like with any, I think with any thing that you start, it does eventually come to, you know, its conclusion. So I think we're gonna see that as well. And probably heading to a different direction afterwards.
Finola Howard
Very cool. Wishing you all the best. Phasma thank you so much for sharing your story, sharing your truth with us.
Speaker 1
Oh, thank you so much for having me. It was lovely chatting with you.
Finola Howard
I hope you enjoy that episode. And if you'd like to find out more about Fatima, connect with her on LinkedIn and check out finised.com That's f i n i t e.com. And if you'd like to support the show, follow or subscribe on your chosen platform. It really does make a difference. And reach out and let me know your takeaways from this episode. What would you like to know more about? I'm looking forward to getting your message out
Transcribed by https://otter.ai