How mature do you think online clothes shopping is right now? You’d think it’s pretty advanced, until you run into the problem of getting your size right. That’s right! A lot of online shoppers struggle with finding the right clothes for their body measurements, and fashion companies having 10 different ideas of small, medium and large doesn’t help, either. This is what Parallel tries to solve. It uses AI to help shoppers get access to the clothes they want by showing them what someone with their body size can wear. It’s as close as you can get to in-person shopping without lifting your butt from your couch. Chris Schmidt joins us in this episode of Edge of AI to tell us more of their groundbreaking work. He also shares some stuff about people management that he believes is the key to Parallel’s success. Tune in for more!
- Online shopping returns due to incorrect fit are a significant problem, accounting for 40% of eCommerce returns.
- Parallel, a user-generated content platform, aims to solve this issue by allowing users to upload their measurements and photos, facilitating better size choices. The platform’s rapid growth is driven by TikTok and has a global reach. Brands are showing increasing interest in leveraging the data Parallel gathers.
- Parallel employs AI for precise clothing size recommendations based on user-uploaded measurements. The platform utilizes AI across content curation, fraud detection, moderation, and recommendations. It also enables users to monetize their content through brand affiliate partnerships.
- The founder’s commitment to focusing exclusively on developing Parallel and resisting the temptation to diversify into other AI domains or industries, highlighting the importance of staying focused in the early stages of a startup.
- The discussion of AI’s role in analyzing user-generated content, such as images, to improve product recommendations and understand user emotions and preferences. This underscores the evolving complexity of search queries and the need for AI to fill data gaps to provide more accurate results.
- “40% of online purchases are returned, with 60% of those returns attributed to incorrect fit.”
- “Parallel’s user-generated content showcases customers’ outfits, resembling professional modeling shots and driving engagement on the platform.”
- “The accuracy of the system is incredibly high. It’s 98% accurate compared with a 3D body scanner.”
- “The fascinating component about Parallel is we’ve essentially lowered the bar for anyone that wants to be a creator. Anyone who has a photo of themselves can become a creator on Parallel. You can be an entrepreneur.”
Listen to the podcast here
Social Fashion Driven By AI With Chris Schmidt Of Parallel
I am Ron Levy, welcome aboard all. I’ll be your captain for this exhilarating voyage to the Edge of AI and like most of you, I embrace the spirit of exploration and entrepreneurship throughout my life. From starting my own business before graduating high school to traversing the world’s most challenging terrains, I’ve always sought out new frontiers and adventures.
I’ve conquered legal battles, built award-winning homes, and now I lead a public company dedicated to pushing tech boundaries and unlocking our full potential. Together, we’ll navigate uncharted territories in AI. The guiding star in this quest is going to be to ask great questions and I’ll do my best to do that. Buckle up and get ready to embark on an amazing adventure. Let’s set sail. For this episode, Chris Schmidt of Parallel, the social commerce platform makes your fashion explorations shareable. I’m going to tell you a little bit about Chris but this is going to be a super exciting episode.
Chris founded the World’s first Snapchat filter company, Geofilter Studio. Geofilter Studio exploded at an unprecedented growth rate of 24,000% over 12 months and it grew to 85 employees in the first year. It was Canada’s fastest-growing company in 2017 and created over 100,000 filters that accumulated 5 billion views worldwide. Those are all astounding numbers.
In 2018, Chris started Pluto Ventures, an artificial intelligence company with a world-class team backed by a strong group of investors and advisors. Together, they’re building Parallel, a social commerce platform with users spanning 191 countries. Parallel is the fastest-growing social commerce community for those who never know what size to buy when shopping online and want to find the brands of the coolest outfits on Instagram. Let’s start with a little bit of background. Chris, when did you get into AI? How did it lead to Parallel? What was that journey like?
Thank you for having me. I’m super excited to be on this episode. I taught myself to be a software engineer many years ago. I went to university. I was supposed to go into medicine but I decided to start businesses instead. A few years ago, when I started Pluto Ventures, I was seeing the early applications of AI and that fascinated me. I was interested in learning the AI side of software engineering and that propelled me into the AI world.
That was very early on. We should talk about where you’re from because I heard how you said about if I want to correct it. Talk to me about where you’re sitting now and where you’re from.
I’m based in Winnipeg, Manitoba, which is in the middle of Canada. I started my first company many years ago when I was ten years old. That company is still operating. It’s a tree service company that protects Winnipeg’s trees against Dutch Elm disease. Over the last few years, I’ve been building technology companies focused on consumer technology companies.
You’ve been doing it very successfully because I do know you were able to exit your first company. You grew it so fast and so large that it was time to exit and took those lessons. Now you’ve got Pluto Ventures and from that, Parallel. Let’s get into Parallel a little bit. What are some of the problems with online shopping that Parallel is addressing?
The biggest problem when shopping online is you don’t know what size to buy when you’re looking to buy a new pair of pants or a new shirt. There have been thousands of people trying to solve this problem. In terms of the scale of it, some of the metrics we have are 40% of online purchases through eCommerce stores are returned and 60% of those returns are due to wrong fit.
This is a massive problem, both from a financial perspective for brands. What people don’t realize is when you return an item, more often than not, that item is not being restocked and sold to someone else. It’s going to a landfill. There’s environmental play here that we’re also trying to solve. Our primary thesis is if we can help people find the right size when shopping online, we can increase that customer satisfaction, decrease waste and solve the biggest problem.
I’ve heard those stats before on the amount of returns and things. It’s baked into the companies now but it’s monstrous. To me, being able to solve that problem can double the bottom line. The positive bottom line of about any company has got to be incredibly huge.
Free returns are killing companies. You’re starting to see now companies where they’re imposing even a $10 return fee or they’re shorting the window. Instead of having 30 days to return, you have 14 days to return. They’re doing whatever it takes to potentially decrease those returns because that’s killing the bottom-line numbers.Free returns are killing companies. Click To Tweet
I’m imagining you sitting in a chair saying, “I’m going to fix that problem.” That’s a big undertaking. Did it start with that global sentence that I said or was it one piece at a time until you got there?
That was the initial thought. In every company I build, I want to solve a bigger problem than the prior company. The way that I view it is, in my lifetime, if every company I build takes between 5 to 10 years, I can only solve 6 to 7 problems. If I’m lucky, I can work until I’m about 90 if my wife allows me to. As an entrepreneur, if you only have 6 or 7 chances to solve a problem, you’re going to pick pretty big problems. That’s how I view it.
When you say it, it’s so simple and natural but trust me, a lot of people are reading like, “I want to be that guy. I want to go do that.” Hopefully, some of this story you’re going to share with us will help pave the way for others to do exactly that because it’s a big statement. How valuable is that? It’s fantastic. If you talk about all the sizes and styles that are available and all the manufacturers here or there, how will you guys affect making clothes more available to more people in a better way? How does that net effect happen?
The thesis is around Parallel, which is the consumer-facing product we’re building out of Pluto, is this idea of when you’re wanting to know what size to buy. The best way to find out is to see someone with the same body dimensions as you, so the same height, weight, chest size, same and seam and see what they’re wearing and the size that they’re wearing from that brand.
The issue with brands is this. Everyone has heard of vanity sizing where instead of having extra small, small, medium, and large and following that standard, they keep adding extra small, extra extra small, and extra extra extra small. It’s this mindset of a consumer that if you’re a 21-year-old female, you don’t want to be wearing a medium. You want to be wearing an extra small because it sounds better. Sizing has been skewed so much that size guides are completely useless. Our thesis is we’re going to show you someone with the same body size as you. You’re going to see what they’re wearing, the size that they’re wearing, then buy that same item if you want that fit.
We’re going to get into how you use AI to measure bodies. Before we go there, is it doing what you described only from people that have subscribed to Parallel and you’ve been able to get their body sizes or somehow, is it also happening through people that have not?
The fascinating thing about Parallel is all the content on the site is user generated. All the photos, measurements, and data are all being uploaded by our users and all the measurements are also being uploaded by our users. We started out with an obsessive user base. They began to tell their friends and we’re now in 191 countries. It’s moving along quickly but we are relying on our amazing users to tell other people to come to the site.
How many users do you have now? Is that a piece of information you can share here?
We’re sitting at about 130,000. We’ve spent no money on marketing. We’re a startup. We don’t have a marketing budget, so it’s all been a growth hack. We’ve been leveraging TikTok, which has about 650,000 followers on our account. We get about a million views a day on our TikTok account. That drives users from TikTok to our Parallel website, then a percentage signs up and they become users on the site. We have an app as well.
If I looked at a heat map on that, which continents are you prominently in?
The funny thing is we almost have an even distribution around all the continents around the world. If you think of North America, Europe, Asia, and Africa, we have sizable user bases in all those countries. That shows, when we lean into diversity and embrace diversity in different sizes around the world, Parallel is a product where you could be in Taiwan, Tokyo, Winnipeg, or LA and you can all use the site. There are no boundaries. It’s a global site. I find that fascinating. The site’s viewed in about 150 languages. We’re very international.
As you were describing the sizing changes that have happened over the last decades, there’s no universal sizing, to my knowledge. I don’t know much about that industry but there’s no universal sizing because no matter what brand I go to, you can’t take reference points from whatever you knew because it’s different. One hundred thirty thousand is an incredible start. Can I ask you when you got your first?
We launched the app summer of 2023 but in 2022, we only had about 10,000 to 15,000. It’s pretty good. We’re getting there.
It certainly is because this will be a snowball. It’s going to be harder to get to 130,000 than it will be to get 1,300,000. It builds up. No doubt it’s good. As you get a bigger library, I’ll call it, of people, there are some obvious next plays there. Maybe you can speak to them a little bit. The brands are going to want to come and hug you. How does that interaction potentially work?
Brands are funny because until you get big enough, they don’t care about you but when you get big enough, then they love you and want to be part of everything you’re doing. We’re getting to that tipping point with a lot of brands. For example, Aritzia. We have thousands and thousands of posts on our site of people around the world wearing Aritzia. They’ve had millions of impressions on our site.Until you get big enough, they don’t care about you. But when you do get big enough, then they love you and want to be part of everything you’re doing. Click To Tweet
We have all this data about their users, waist size, seam, and everything. We’re getting to that point where we should honestly be calling Aritzia and saying, “Let’s do something,” but months ago, when we only had 200 posts, they wouldn’t care about us. It’s just a tipping point that we’ll have to get over then the brands will start to fall in place.
They’ll fall in place for you but then they can also change their model a little bit and start making clothes because you have one particular size that’s most common. They not only know the sizes from what I understand. They know what they bought, what color, what style, and all of that so you can reverse marketing.
The interesting data point is we can tell Aritzia the average waist size of someone between the age of 25 and 30 wearing their latest pair of pants in Winnipeg compared to someone in Vancouver, England, or London and this is how it’s different. When you go a little bit deeper into the data, we can then say the users that are buying your latest pair of pants are big fans of Converse because they’re tagging Converse on our site as well.
If you’re Aritzia and you’re looking at expanding your marketing capacity and finding a new segment, we can tell you to go after the Converse-wearing people in North America because that’s your clientele as well. The data are quite interesting and if anyone knows any brands and wants to introduce us, we’re happy to start talking.
You are also a data company, without any doubt. A very valuable one. I don’t know what the percentage of returns is or if it’s 50%, it’s substantial, or whatever it is. If I can eliminate that and pick up new sales, I’ve done wonderful things for my company so I will buy your data. That’s where it’s going to go and that’s brilliant. There are a few places in what you’ve been saying where I can imagine AI coming into play and being used here. I want to go back a little bit because you started in AI in 2015. Is that accurate?
It’s a few years ago now. Maybe we can talk about where you’ve applied it in the past. We know a little bit about that but what aspects of this business now, Parallel specifically, are you using AI in? It sounds like there are multiple spots for it.
That’s correct. The core AI we started building years ago is that body measurement technology. What I can do is I’ll show you how we use that and I can talk about the other places that we’re using AI on the site. For those reading, we’re on the homepage of the site. We also have an app. This is a place where you can see other people that have uploaded their photos and have tagged their outfits and the items that they’re wearing. It’s like an Instagram-Pinterest feel and a feed that’s curated to what you want and your shopping preferences.
I want to say these are all modeling shots. These are your customers putting their images out here. If you told me you were a brand and you paid to get all these shots, I would’ve believed you. This is powerful in itself.
This is our users posting what they’re wearing and tagging those outfits. It’s quite amazing. I’m going to jump into it. You can sign up for Parallel. It’s completely free. It’s like any other social media platform. In your account, you have the ability to upload your measurements.
JoinParallel.io is what I’m seeing. Is that accurate?
That’s right. It’s on the iOS store and it’s coming to Android as well. In your account, you have the option if you want to upload two photos of yourself. A front-facing and a side-facing photo. I’m going to upload those from my system and a few comments about the photos. The accuracy of the system is incredibly high. It’s 98% accurate compared with a 3D body scanner.
Wearing tight-fitting clothing does help the accuracy. We realize that not everyone is comfortable wearing skin-tight clothing. We built an AI module on top of our system that will accept more baggy clothing. A baggy T-shirt, a hoodie, or jeans, and the AI will infer your body measurements under that baggy clothing. We wanted to build a system that everyone felt comfortable using the pose.
That’s a good point. They’re sending their pictures or images to a stranger. The screen we’re looking at has someone that looks like you but is not you.
This is me.
That is you. It looks like a Black body suit, which probably is not. It’s tight-fitting from top to bottom long sleeve.
That’s right. The pose I’m standing in is called a pose on the front side then I’m standing with my arms by my side on the side shot. We have very strict data privacy laws that we follow and we are very careful with users’ data so we don’t save any of these photos. You enter your sex and you put your height in. What will happen is you hit submit. In a few seconds, we analyze those photos. We capture thousands of body dimensions behind the scenes. Chest, sleeve, waist, seam, and hips are the five core metrics that you need when you’re shopping. We generate those and we attach those measurements to your body. These measurements are so accurate. I can take out a tape measure. I can measure my body and they’re pretty much spot on. I’m going to jump into our shopping page.
We went over to a different page. We had choices between shirts or pants or anything else. You clicked on shirts and it looks like it’s all men. You click male, so we’re seeing all men now.
I’m going to click on a post. Here’s Jonah. I can compare his body measurements to my body measurements. You can see across height, chest, sleeve, hips, waist, and seam that we’re pretty similar in measurements. We’re off by about an inch on some of these measurements but what that tells me is how clothing fits Jonah will fit me in a very similar way. Jonah has linked up his items and I can click through right to the retailer. I can buy that shirt, watch, or pair of pants.
I’m going to go backward a little bit. There were two rows there. One was Jonah’s and one was Chris’s and it showed the five measurements. It might have been 5 or 6 measurements. Three six of them, height 5’10”, height 5’10”, chest 38 and chest 40, sleeve 23 and sleeve 22. They’re all very close. You can take that into account but you can see what it is. From there, you click on what?
If I want to buy the same T-shirt that Jonah is wearing, he’s linked it up. I would hit Buy Now. This is going to take me to the RW website. This shirt is out of stock, so let me go back because he posted this a few months. What I’m going to do is I’m going to jump into Jonas’s profile. He uploaded a Jimi Hendrix T-shirt. He is wearing a size large T-shirt. I’m going to click the Buy Now if I want to wear the same T-shirt. He’s linked it up and I can go right through and buy this T-shirt. It’s on sale. It’s $10 and I would buy a size large because that’s what’s Jonah is wearing.
The ship to the credit card or however they’re paying and all that is built into Parallel, correct?
At the moment, we pass you off completely to the retailer. All the purchasing is still taken care of by the retailer. We’re simply a middleman helping you find your next outfit.
There was a company, a number of years ago now, they were on Shark Tank at a certain point, used AI. They used your own cell phone to take some photos but the difference was they manufactured the clothes. I don’t know if they’re around now. They are doing quite well but I haven’t heard of them in a few years. They’re manufacturing clothes. What you’ve done here is a different play. It’s purely tech play. Do you want to define it in a different way?
That’s a good way to put it.
It’s scalable and expandable to even more countries than you’re in now and more markets. That’s pretty amazing.
We work with thousands of brands. Any brand on the internet, if someone buys your product and links to it through our site, people can shop it. There are no limitations on scale and opportunity. It’s every brand in the world and every user. People ask me, “Who’s Parallel for?” The answer is, “Anyone that wears clothing can use Parallel.”
The original question was, what aspects are you using AI in? The measurements are what we focused on, so it’s amazing for that. You’ve got that nailed. Are there any other parts of the business you’re using AI in?
People might be surprised at how many AI components a platform like Parallel has. All those recommendation feeds, when we’re serving content for you to buy, is all AI generated based on what you’ve been looking at, the brands you’ve been wearing, and the price points you’re buying. All those feeds on the site are curated to you very much like Instagram or TikTok.
On the backend, we’re using AI for fraud detection and where abuse might be happening on the site. We’re using AI to leverage that technology in those areas. Also, analysis of photos and moderation. All the content coming into our site, we moderate it. We want to have a site that’s usable and that you can shop from. When people upload pictures of their cat or their toaster, that’s not quite the place for our sites. We use AI for content moderation. Almost in every asset of our business, we are using AI behind the scenes.
Talk to me about opportunities for entrepreneurship for those people outside of Parallel that want to get involved. I think you’ve got some opportunities there.
The fascinating component about Parallel is we’ve essentially lowered the bar for anyone that wants to be a creator. Anyone that has the photo themself can become a creator on Parallel. You can be an entrepreneur. We have some people that after school, they go to the mall. They’re taking photos in Aritzia’s changing room. They leave the store and haven’t bought a thing. They go home and they upload everything to Parallel.
You now have someone who’s an Aritzia model and they’re beginning to build a business around that. The way that our site is set up is we have affiliate partnerships with thousands of brands. When you upload a photo to Parallel and someone buys through your post, we get a commission back from the brand. We give that commission to our creators. A lot of our creators are now becoming models and monetizing their outfits on Parallel. It’s quite an interesting loop. Once again, if you’re wearing clothing and have a photo of yourself, you can become an entrepreneur and a creator on Parallel.As long as you’re wearing clothing and have a photo of yourself, you can become an entrepreneur and a creator on Parallel. Click To Tweet
That’s brilliant. It’s fantastic. They’re earning cash credits. Is that what they’re earning? Is there a tokenization on this or is it credits with Parallel versus credits with the individual manufacturer? Do they typically get end up with some cash cheque or do they get credits and be able to buy their clothes? All of those questions.
At the moment, you get cash into your PayPal account. In the future, we’re looking at opportunities to do tokenization or purchasing within brands. We only started building Parallel about a few years ago so it’s very fresh. At the moment, we will send you PayPal money.
Look how much you’ve done in a few years. That’s an amazing build.
I have an amazing team. There are thirteen of us. I’m the dumbest person in my company, which is amazing because I only hire people smarter than me, which is easy to do. It’s an amazing team of full-stack engineers, machine learning engineers, marketing, biz dev, and growth. They’re the brains behind the actual site. I’m the puppet at the front showing off what they do.
It sounded like you mentioned affiliate marketing there. You are already generating some revenue. Is that accurate?
We are. One of our investors helped sell Honey, which is the Chrome extension that helped you find coupons when purchasing online. She helped that sale from Honey to PayPal for $4 billion. She’s one of our investors on the team. She has extensive knowledge of affiliate networks and monetization. I’ve brought onto my team and my investor side people with very deep knowledge in the domains that we’re playing in. Affiliate marketing is a big space that we’re in.
I’m curious about you. What are your days filled with? Is it actual coding and tech designing or running the business and doing other things? What does your typical day look like?
Going back to my thesis of only hiring people smarter than me, I’ve been good. I have not pulled my repo. I don’t have the repo on my site, so I’m not doing any coding. To satisfy that, I’ll sit beside my developers and I’ll bug them as they’re writing code. I’ll be watching their lines and giving them suggestions.
I’ll realize I should probably go do something that I’m supposed to be doing. At the moment, we’re wrapping up our seed rounds. I’ve done a ton of investment raising over the last twelve months. I lead all my teams, whether it’s the development, marketing, or growth team. I’m planning out all their schedules and the targets that we want to get to. I’m very deep.
I’m deep in the data and the KPIs. I’m working with the team. I’m having very deep meetings with the guys but I’m also doing corporate taxes and legal. I’m the definition of wearing many hats at the business. Every day is much different than the last. At the end of the day, I’m keeping the glue of the company together as it keeps going down the pipeline.
It sounds like you are working out of a physical office. Is that correct?
That’s right. We have an office. On Fridays, we work from home. It’s one day of the week that the guys in the summer work from home. They usually get off early in the afternoon. We’ve been in the office the whole time. Even through COVID, we had the right protocols in place. If anyone felt sick, you would test. You would stay home and get checked. We worked in the office all through COVID and we had no COVID cases for two and a half years. When you have the right measures in place and you have people that you trust, you can work around these issues.
All thirteen of you come into the office. Is that accurate?
That’s right. The office is what you would imagine a tech company is. There are dogs, beer on tap, and Red Bull in the fridge but on the flip side, everyone is working hard. Some nights, we’ve been there until midnight to 2:00 AM. We have deadlines. It’s a tech startup and if you don’t make enough progress, you’re going to run out of money. There is pressure to get stuff done. It’s not just fun and good times every day but we make sure we have some fun occasionally.
When you said that you started this a few years ago, it was amazing to me because of that level of build, what it took, designing the company, and getting your investors, board, and advisors. That for 24 months is amazing. I’m wondering, could you have done that if you were all virtual? Do you think that being in a physical space with all thirteen of you made a substantial difference?
It’s the reason we have been successful and is the reason we will be successful going forward. There’s a special spark you have when you’re in the same space. That spark people that start a company will understand this magical spark where every day is hard. Every day, you have challenges of people hacking the website or people trying to game the system.
You have all these difficult situations to solve. When you’re there in the same place with everyone, it makes it a little bit easier to get through those days. I’ve built a culture. For me, culture is everything. It’s the most important thing. When you build a culture where people want to be with other teammates in the same place, it allows you to have a little bit more productive output and you can work a little bit harder. You can push those 8-hour to 9-hour days because the guys want to be there for nine hours.
It’s not me forcing them to be there for nine hours. That magical spark, you can’t replicate over Zoom. If you’re a massive company, have remote people but if you’re thirteen people, the speed we move at, you can’t match. In the early days, we were pushing to production, so deploying new code to the website 6 or 7 times a day. Our iteration cycles are very fast. The speed we move at, we have to. We’ll get eaten alive if we don’t.
I concur completely. There’s one aspect that you didn’t bring up but it’s very relevant. That is the creative side. Virtual, you can do tasks. Each person has their tasks. They’ll do their tasks and you’ll check in. You’ll spend a minute or an hour together and say, “We did this task,” and away you go. Being in a physical room, especially with a startup where creativity is important, is collaborative creativity.
It is those conversations that you think are going to be nothing. They’re not even important and all of a sudden, some gold comes out of it. I would say, back in the day, that used to happen in a happy hour after everybody left the office. The same thing. There were certain magics that came from it. That, to me, you cannot get that virtually. It’s something we’re all struggling with. How do you do that dance? It sounds like you nailed it.
It’s more often than people realize that the intern or the junior developer makes a side comment in the office. That turns into being spot on and you implement that idea right away. I agree with you. In the very focused Zoom meetings, you don’t get that very natural conversation and idea generation. I’m all for being in person.
We’re going to back and focus on AI a little bit. AI is going to be shaping all industries, most products, and most services, even. How do you think that, moving forward, AI is going to be shaping the field of social commerce, which is what you’ve endeavored into?
The challenge in our online domain, especially with shopping, is there are so many options. There are more brands than ever now. Shopify makes it so easy for anyone to start a brand. There’s consumer overload on, “What should I buy? What’s good quality? What’s not good quality?” This is where Parallel is fitting into this nicely.
Instead of searching Google black T-shirt, the recommendation and the filters are not tailored for a good shopping experience. There’s an opportunity. This is what we see that we can fill that void of an easy and simple shopping experience where you’re not overloaded with 10 million products. However, because we know you and we know your body size and what you like and different brands, we can say, “Here are four T-shirts across four different brands that we think you would like. Here’s the price point. They’ve been on sale for two weeks. Feel free to buy them if you want.” I feel like AI is going to simplify the shopping process because now, you can be shopping on Instagram, TikTok, Pinterest, and on the brand’s real sites. There’s so much and simplicity is important for a consumer.AI is going to simplify the shopping process. Simplicity is important for the consumer. Click To Tweet
Agreed. Parallel has grown so fast. No doubt you have a roadmap without goals. You flounder. You obviously are good at setting goals. Where do you see Parallel by the end of 2024?
I would love to get to a million users. It’s a vanity benchmark that allows you to show up on the scene and say, “Hey.” It’s the same issue you have with brands where when you’re small, no one takes you seriously. A million users is a vanity metric that where people say, “This seems to be working.” It’s been working for many years before that. That’s a vanity goal we want to get to.
At the end of the day, we’re trying to build an environment where no matter what body size you are, you feel comfortable posting on Parallel. We have individuals that were 3 extra large and 4 extra large sizing. We’ve built a community where they feel comfortable posting on Parallel. They don’t feel comfortable posting on Instagram. They don’t feel comfortable posting on like to know. We want to build an environment where people feel safe, can shop and find what they want in one place, and get to a million users at the end of the day then we can keep moving up after that.
I’ll go back to my earlier comment. You’re 13% of the way there. That first 13% is certainly the most difficult. I’d throw down a nickel and say that you’re going to get that well before the end of 2024. I completely concur. Pardon the pun a little bit but with Parallel, you have Pluto Ventures. Are you paralleling at this time and doing other projects? I’m sure you will be at a certain point but how are you doing over there at Pluto Ventures outside of Parallel?
Parallel was developed three and a half years after Pluto started. The initial thought with the technology was we would build an API for enterprise integration into a variety of industries, so healthcare, insurance, and commerce. We would be a backend support tool that provides value to the customer on body measurements.
What we found was we were honestly too early for a lot of those industries at the time. This was back in 2018, 2019, and 2020. We focus 100% on Parallel. When I build a company, I do believe focusing on a niche is the best option to be successful. A lot of people go too wide too soon then they get resources stretched and don’t work. We are 100% focused on Parallel but we do have the opportunity in the future coming up to use our AI body measurement system as an API call for other systems but that’s not a focus. My investors like what I’m focused and we’re going to stay focused on Parallel.
Focusing on AI, are there other AI domains you’re in now, or the answer you gave packaged it? It sounds like your focus is Parallel now and that’s what you’re doing. It seems like the right thing to do, by the way, based on what you’ve described.
We see a lot of opportunities. Continuing to dive into the vision AI sphere for the body measurements, we’re doing image analysis at the end of the day and extracting body measurements from the photos. As you keep going down that path when you think about shopping and people uploading photos of what they’re wearing, there are lots of opportunities to use AI for helping to detect products, photos, colors, and patterns and to simplify the user experience on the creator side.
That’s where we’re heading. There’s a lot to be done on that side and even understanding if the person is happy in a photo. Instagram has been doing this for many years. Vibes, feelings, and emotions can be extracted from a photo. If someone is searching on Parallel for fun summer dresses, using the word fun is a very difficult thing to bring up search results for because no one is tagging fun when they’re uploading a photo.
You can extract fun from the scenery, the person, or the emotion from that photo. There’s a lot that can be done because what we’re seeing from users is that you’re not searching for a very simple black Jimi Hendrix T-shirt. You’re searching for what can I wear to go to a concert? That’s where the complexity of a search is evolving too. You have to use AI to fill the void of data to provide the results for those more complicated search results.
Parallel is going to be busy for many years to come. You showed how many years of work there is to do and loving every day as you’re going through it is amazing. You’re super great. We’re going to shift a little bit and we’re going to go into AI wants to know. This is where we get to know you a little bit better. AI, by nature, is built on curiosity, which we all are curious or we wouldn’t be here. These are ten quick questions designed to uncover intriguing mysteries that AI longs to comprehend but can’t quite grasp. Think of it as a snack break in the journey. Keep the answers quick but the safety belt sign is off. Let’s explore more about who you are and what makes you tick. Are you ready for this?
What’s the first thing you ever remember being proud of?
Cutting my first lawn when I was six years old.
What do you need help with that you wished you didn’t?
I’m probably not located in the right city to be building a tech company. Winnipeg doesn’t have an abundance amount of tech companies around. I wish that I was in a city where I could walk down the street like in San Francisco, LA, Boston, or wherever it might be, and there are twenty tech companies on that street. There might be ten tech companies in all of Winnipeg. It’s a pretty small ecosystem here.
I’m sorry, I got to drop a comment. It’s why you’re so focused. Those are distractions. I bet Winnipeg is a perfect answer. You have thirteen perfect people building an amazing company. I wouldn’t know what to look for elsewhere. What do others often look to you for or help with?
I’ve now started a few companies. I have a good understanding of how to go from 0 to 1, from idea to whatever one people believe is, which is the first MVP, product market fit, or whatever that might be. In that early stage, which is usually pretty hectic and hard, I feel like I have a pretty good path to walk through that and not drive myself insane.
What do you treasure most about your human abilities?
I have a high level of grit. I can withstand a lot of stress and uneasy situations and coast on through. I don’t get mad. I don’t yell. I am neutral the whole way through. My emotions don’t swing up and down. From a vanity perspective, I like my hair. I’m thankful I have a full head of hair still, so that’s a weird two sets.
I’m with you there.
You as well.
My dad kept all of his all the way into his 90s.
I don’t take it for granted.
That’s right. What is the most consistent thing about you throughout your whole life?
My temper and my drive for working and being successful. As I said, I was working when I was 5 or 6 cutting lawns with my brothers, and 25 years later, I’m still working eighteen-hour days on a Friday. I like to work.
We know it stayed the same but number six is, what’s changed the most throughout your life?
My ability to assess a situation, not make an impulse decision, and being able to understand. We’re talking a little bit before the segment. When you have an employee that maybe comes in late or seems not to be working to the full capacity that they can be, do not assume anything but instead ask them how are you doing. “Is there anything I can help with?” Taking a step back and not always assuming the worst that they’re trying to not do any work and they’re trying to leave the company stuff. Patience, taking a step back, and always assessing a situation, maybe twice.Have patience, take a step back, and always assess the situation, maybe twice. Click To Tweet
It’s something that a lot of people either learn too late or don’t learn at all. The way I try and sum that up is to be the question. Not the answer. What do you find strangest about reality?
I have a weird obsession with space. I have a science background. The concept of space is fascinating. The size of it is fascinating. When you do think about how small you are in space and sometimes things are making you upset, you have no impact and made it be you being, for lack of better words, dumb at the end of the day. Understanding the perspective is important and realizing how big space and time is mind-boggling.
Makes plenty of sense. When most recently do you remember feeling alive? That feeling we’ve all gotten at times when you’re feeling the most alive.
Throughout the journey of a business, there are all these moments I always talk about that are so thrilling. For Parallel, the first hundred users were fascinating. The first time we got to 1,000 users, the first time we had so many users, our servers overloaded on AWS and the whole site crashed. Those moments. It’s almost in moments of chaos. I feel alive because it shows that it’s working. If you have enough people breaking your site, it’s working and good. We broke all of our servers many times and that was exciting.
There’s a question I should ask you earlier and I’m going to hit it now. How did you get your first twenty?
Manually messaging people and begging them to sign up. The misconception is when you launch a product, people think that you will post it on social media and a million people will flock to you and everything is going to be fine. The reality is you will have to convince your parents to sign up and you’ll have to take their phone and say, “I’m going to sign up for you.” For the first twenty users, we basically begged, “Please sign up.” It was the first 500 users we begged to sign up. All it takes is one of them to say, “This is cool. I’m going to tell some of my friends,” and then ten more people come on. That’s how you keep going.
That makes complete sense. It is completely aligned. By the way, that was question number eight. You’re almost out of the hot seat but what’s your most unique trait?
I have a strong ability to build a sense to-do list and to execute and to be doing ten things at once all in sequence. Still, for those ten different things, I’ll be doing 100% focus on every single one of them. My ability to be managing thirteen people in sequence at the same time and all different projects to all align with the same end goal, which is 6 to 12 months out. I’m starting to realize it is a unique ability. The way I operate is very natural to me to do that. I don’t have any stress waking up trying to manage thirteen people smarter than myself.
This is your last question here. If you weren’t human, what would you be?
I have always had a fascinating interest in tigers, monkeys, and penguins. It’s weird. Curiosity is another strong trait of mine. Monkeys have a very high level of curiosity as well. Curious George is always a favorite book of mine growing up. I’m probably a monkey.
I get it. I did spend some time around some chimps and they are that curious about everything. Don’t turn your back.
They’re very curious.
We’re going to head into the next segment. We’ll talk about AI leaders and influences. Go ahead and highlight some of the leading individuals, projects, and organizations that influence you or that people might want to be interested in following.
There’s the obvious Elon Musk and Sam Altman. Elon Musk is running a million companies now, it seems. Everything from SpaceX to Tesla to X. As of now, it’s not Twitter. It’s X, then all his other companies. Sam Altman, who was the President of Y Combinator for many years and now is leading OpenAI, which is the company behind ChatGPT, which almost everyone now in the world has heard about.
Those two are very prominent, in your face, in the press everyday AI leaders that if you’re an AI, you can’t not watch them, follow them, and see what they’re working on. What’s interesting about AI is that there are so many people behind the scenes that you don’t realize. Software engineers and developers, by nature of them, are very quiet and typically introverted type people who aren’t looking for accolades and being on the front page of the paper. There are a lot of very quiet people behind the scenes that are moving the ball along. I’m very fortunate. One of my investors is considered the Godfather of AI. He’s based in Toronto. He doesn’t want his name mentioned in the press. He has a few books out there that he’s written about AI.
He’s someone who, when you think about the last few years, he’s moved the ball of AI along behind the scenes. You have to keep in mind that Elon Musk himself is not building neural networks at night. There’s some 25-year-old engineer sitting in a Winnipeg office tweaking nodes and data sets. The progression of AI will come from that guy.Elon Musk himself is not building neural networks at night. There’s some 25-year-old engineer sitting in a Winnipeg office tweaking nodes and data sets. The progression of AI will come from that guy. Click To Tweet
My hat is off to the Godfather of AI you described. He’s obviously done all that he’s done egoless and that’s so powerful in itself. The fact that you’d love to shout his name out but you’re respecting him and you don’t says so much about him. I love it. Who’s influenced you the most, both within the world of AI and maybe outside too? I was going to say business in general but it doesn’t have to be. This is when we want to recognize someone that’s had an impact, whether it was a direct personal impact or through the work that they’ve done.
There are specifically three senior machine learning engineers at my company that started with me at the beginning. It’s Yun, Jesse, and Dami. When I hired them, as I mentioned before, I was self-taught on the full stack development side but I was quite lacking on the AI side. I wanted to build an AI technology company and I knew that I had to get myself up to speed. One so I could talk to these guys that know way more about me but how will I be able to lead them if I don’t understand their world and terminology?
For those three guys specifically, I would ask the stupidest questions a million times over and over again with them. I’d be like, “Why are you doing that? What’s that epoch? What’s overfitting?” I kept asking. Technically, I was their boss. They were working with me but they were so patient. They would explain in extensive detail what they were doing.
Quite honestly, they were my teachers. I would go away, read articles, watch YouTube videos. I’d come back to the office the next day and be like, “I saw this. I read this. What does that mean?” I sought out people that, quite honestly, could teach me and I built my team around that. On the other side of the business side, at the end of the day, my parents and my family are massive supporters of what I do.
None of them had a business background. My dad is a physician. My mom is a trained architect but has been a stay-at-home mom for many years. There was no way for them to support me from a business perspective but they supported me in other ways. People always joked about this but my mom would make me lunch when I would go to work when I was 27.
People would laugh about that. I would say, “Yes, but my mom making my lunch gave me ten minutes more that day to be writing emails.” That was how she was able to support me. That is people don’t appreciate that you may not have someone that’s directly supporting you but coming home to a warm dinner from your wife is support and that’s a very different type of support.
It sounds like your parents may be another great reason to stay in Winnipeg.
That’s why I’m here.
Have we missed anybody you want to give a shout-out to that deserves it in your world or you could think of that we haven’t handled? Certainly, if you want to share their Twitter handle or anything else.
Some of my early investors, Ty, Ivan, Amir, and Sam invested when it was an idea a few years ago. It was purely, “You have this crazy idea, Chris. We want to help you. We’re going to give some investment.” Along the way, all my other investors have come on board across the whole spectrum has been a lot of support.
There’s a lot of support in Winnipeg. There are a lot of organizations that have been helpful. I got married in 2022. I’ve been with my wife now. We’ve been together for nine and a half years. Leah has put up with my crap for years. When we started dating, I was writing the MCAT. I was going into medicine. That did not happen. Now, I’m here. For years, she had been so supportive. I’ve been home late by like six hours. I’ve been totally off on things and she’s never complained. She’s always there, supportive, and willing. When I come home, she give me a hug. No matter what happened at work that day, it doesn’t matter. It’s like she wants me to be home. She’s amazing. That’s why I married her.
Congratulations on being a newlywed. It sounds fantastic. There are three levels that could go and you describe the helpful side. You are with someone the way you work. She’s making your days better and more productive in all those ways. You can also have relationships where it’s neutral. Nothing negative but it’s neutral but you’re not getting the extra support. We can also have those relationships where it’s taking the energy away.
It sounds like you’ve nailed this. Congratulations to you for her and her for you. It’s pretty amazing. Let’s shift to the AI resource list. This is where I’m going to ask you to share a handful of your favorite resources in AI. Think about all of our readers that are in the world or wanting to get in the world. It could be websites, applications, books, podcasts, or learning tools. If there are any of those, you can share them to help their journey.
All my software machine learning engineers have gone through the deep learning Stanford online course. CS231N is the actual course number. There’s a whole bunch of lectures. They’re about an hour or an hour twenty lectures. There are about 15 to 16 of them. They basically take you through the course of being a newbie in machine learning engineer.
They walk you through all the concepts, different terminology, models, data sets, structures, and everything. They all did this course and it projects you into the AI field. That’s a good resource educational piece. I do have to say having some foundation to do when you do that course is beneficial. Having a Computer Science degree, Computer Engineering degree, or being self-taught on the coding side is going to be beneficial. It’s not just a course I would recommend a business student takes.
That’s a very good practical course online. There’s an immense amount of websites these days where you can go to the website. You can begin to play with models in real time. One of those is Playground.TensorFlow.org. It’s this cool website where you can add different nodes. You can adjust the epoch, the data input, and the data output. You can adjust all these settings around a neural network, which is the core component of AI machine learning.
When you add more nodes to a neural network, you can see how that affects the data output. There are lots of hands-on tools these days, whereas a few years ago, there weren’t as many as you would think. You have to have a little bit of that curiosity that a monkey has and search Google for these tools to get you up to speed if you want to dive into AI.
That sounds like a great tool, Playground.TensorFlow.org. It sounds fantastic. How about some AI tips? A lot of us have been playing with different AI programs that the mass markets are utilizing now. Are there any tips you can throw down to us that’ll help the way we’re utilizing them?
Coming from a full stack development background in machine learning, there were a few things that I was naive about that we learned quickly. First of all, the development cycle lifetime of an AI project is much different than, let’s say, you’re building a website. Building a website, there’s almost a clear start and end point and you can launch the website, people use and it’s going to work.
With AI, there’s no finite endpoint. You start the project but the end comes down to what your level of completeness has done and what level of accuracy you want. You may be chasing a 1% of accuracy increase for an extra year or two. It comes down to the question of, does it make sense? That’s my next comment. When you’re striving for perfection, that is difficult in computer science and computer engineering because there are always bugs and issues happening. With a traditional website, you can get to the point of it works and it makes sense.
With AI and I’ve had my engineers have to get better at this. They have gotten better at this. They ask the question, is this good enough? It will drive you insane trying to get that last piece of perfection. You have to have a different mindset. The other thing from an AI perspective and my developers have told me this. You may go 2, 3, 4, 7, 8, or 9 months without getting a win under your belt.
Mentally, that can be very difficult. You could be working for longer than 2 or 3 years on a project and you never have that win where you’re satisfying that human nature that you need to keep going. Whereas, when you’re building a website, if you get that signup button in, that’s a win. If you get that form in, that’s a win. If you get the homepage working, that’s a win.
From a mental perspective and a psychological perspective, if you’re like me running an AI team, you may have to start to realize my team needs faster wins or they’re going to burn themselves out. That’s something that we are very aware of now. The guys went three and a half years with no win until they turned on the engine. It worked and it was an amazing win but took three and a half years to get there. It’s too long.
That makes so much sense. It’s such a big journey. It may not have been too long but it’s too long to go without hitting the champagne cork moment. It’s a long time to wait. I was talking to an artist and I said to him, “I could never be an artist.” I’m talking about a painter, someone that draws and does physical manual dexterity art.
I said, “I can never be an artist because I’ve always got to do one more brushstroke. I’m not going to stop until I ruin it, unfortunately.” His comment was, “That’s why we signed our art. When you put your signature on it, you’re done.” I thought that they even got their closing. Is there anything we haven’t covered here? Does anything come to mind for you? Is anyone you want to mention at all that we haven’t covered?
I want to touch briefly on the fear that people have with AI and where AI is at. Everyone thinks that AI is going to take over your job and everyone is going to be jobless. That’s not the case in a very practical use case of AI. My developers, who I still need, use AI as a superpower. When they get stuck on a problem or a math equation, they need help to get over that.
Traditionally, you would’ve spent maybe half a day or a day trying to solve that one problem. Now, they have ChatGPT. They can ask that one question. They get their answer and keep moving on. All my engineers that have 5, 6, and 7 years of experience now have 20 or 30 years of experience and have this superpower.
What’s interesting and where people are a little hesitant and I agree with this is, at the moment, AI has no emotional side to it. It doesn’t understand incentives and wants. AI is not going to destroy the world tomorrow. There’s a misconception around that. We have to start to become careful when we start to train AI systems to understand emotions, feelings, what humans want, and what the Earth wants. If you can start to manipulate that, that’s where is an interesting component of AI.
Let me give you an example. We know that humans need money to live. It’s a necessity. You need to buy food. If an AI system understood that humans, in a sense, rely on money to survive and you want to cause harm with AI, you may train an AI to understand that if I can take away money from people, I have a way to destroy humanity.
Those are the challenges. If the AI starts to understand the banking institution is where the money is housed, if I can get into that or destroy that, that’s where the next level of AI has to be a little bit more thought about. It’s when you start to have a human perception behind the AI. That’s where people haven’t quite got to yet on the development side.
That is a good point. Control the money, control the people. I hadn’t heard anybody bring that up in this around AI but you’re right. It makes a lot of sense. Chris Schmidt, you’ve been amazing. Where can readers go to learn more about you or follow you and your projects? Also, what’s your ask? What could people do if they’re as impressed as I am with what you’re working on in Parallel? How could they support you?
You can find me across all the platforms, LinkedIn, Instagram, and Facebook if people still use that. Feel free to connect. I’m always happy to connect and chat with anyone. Quite honestly, if you want to sign up for Parallel and tell me what you think, I love user feedback. We crave it. We want people to tell us what they think. That’s pretty much the main ask. Come connect with me on social media and go use Parallel and go find the next outfit.
Why don’t you throw a couple of those handles out there?
Chris Schmidt and my handle on Instagram would be @_ChrisSchmidt. LinkedIn is Chris Schmidt and Facebook is also Chris Schmidt. Those would be the main ones, then Parallel social media, the handles would be @JointParallel.io and that would be across TikTok, Instagram, Facebook, LinkedIn, all those platforms. The website is JoinParallel.io. It’s also in the Apple iOS store. It’s coming to Android soon.
Thanks so much for taking the time with us. We’ve loved every minute of it. It’s time for another safe landing at the outer edges of the AI universe. This is Ron Levy and on behalf of our guest and the entire crew, I’d like to thank you for choosing the voyage with us. We wish you a safe and enjoyable continuation of your journey. When you come back aboard, make sure to bring a friend. Our Starship is always ready for more adventures.
Head over to Spotify or iTunes now. Rate us and share your thoughts. Your support and feedback mean the world to us. Don’t forget to visit EdgeOfAI.co to learn more and connect with us on all the major social platforms by searching for EdgeOf_AI. Join the conversations that are happening online, which you’ll find there. Before we sign off, mark your calendars for our next voyage. We’ll be continuing to unravel the mysteries and advancements of AI. Until then, see you.
- Geofilter Studio
- Pluto Ventures
- Curious George
- Y Combinator
- @_ChrisSchmidt – Instagram
- Chris Schmidt – LinkedIn
- Chris Schmidt – Facebook
- @JointParallel.io – Instagram
- TikTok – Parallel
- Facebook – Parallel
- LinkedIn – Parallel
- Spotify – Edge of AI
- iTunes – Edge of AI