deepPIXEL Startup Pitch - Reducing Business Costs

Uh, now we're joined by Deep Pixel. Welcome. Uh, do you want to quickly introduce yourself? It builds intelligent digital agents for businesses to reduce costs, increase conversion, and serve customers. Welcome on screen.

Yeah, thanks, Adair. Um, yeah, so that was a great introduction. Uh, what we do is we build the AI systems that power the next generation of chatbots. And I'm sure you've heard, and a lot of people have heard of chatbots lately. They're the new hot thing for companies to build and for marketing and for customer engagement. But we're trying to take it one step further and we're trying to automate some support functionality using the chatbots. And with that, we bring in the AI systems.

Excellent. So why don't you do the awkward switch to your presentation and I'll try and entertain everyone while you do that. I have a really great juggling act, but I'll reserve it for later because my office isn't very big. By the way, I don't normally do the webinars in my office. So while you switch, I'll just talk. But this is actually my office. So welcome, everybody. All right, here we go. Excellent. So when you're ready, I'll start the timer. Deep Pixel, everyone.

All right. So yeah, I'm Filipino, not Geary. So I help with the business development and I bring in clients. Some of our clients are very large banks and some other companies are large retail players that are selling products and they get a lot of customer service questions, requests, this isn't working. And we're trying to automate a portion of that and help the customers focus on the more complicated questions so they can spend more time with the users and they don't have to waste as much time waiting on the phone. For their customer service people to pick up, which I'm sure everyone has experienced before. And so here's the problem, right? So companies spend millions annually on customer service. And the issue with that is a lot of times even really wealthy companies with like tens of millions of dollars only have one customer service center in one geography so that they don't have 24-hour service.

—Can you speak a little louder, please? Just a number of people. Thank you.

Or closer to the mic. —Yeah, so there's a really high employee turnover rate in a lot of these customer service centers because generally they're answering the same questions over and over again. And that's what our system is aiming to solve. So what's really the solution? So with Deep Pixel, companies can save over 40% of the cost of having someone responding to live chat requests. And that allows the company, because the system is automated, to have a basic level, a tier 1 support level of customer service 24/7. And a lot of times what we've seen with our clients is they get a pretty quick ROI of about 5x in the first year because we can also drive conversions and drive sales of products through live chat. Now here's, you're seeing the product and these are a few different demos of things that we've built for some of our clients. So on the left you see we built for an airline company a chatbot through Facebook that reduces the amount of time it takes the pilots to schedule their flight times and schedule what planes they'll be flying. In the middle you see something that we've built for one of our banking clients And this allows them to increase conversion for retail banking. So they had actually a simple chat system before and they wanted to upgrade the level and the amount of functionality because they saw a significant increase in the amount of conversions that their website was driving. And on the right here you're seeing reduced costs with this e-commerce bot that we built. So this was actually Beer Bot. It allows you to order beer right through Facebook Messenger and have it delivered directly to you. So now companies don't necessarily need an app anymore. They don't need to force people to download an app. They can just message the company on their Facebook page and then have beer delivered straight to their house. Now what's our business model? We deal with pretty large enterprises usually, so we ask for $50,000+ upfront, and then we also charge in terms of like support and maintenance depending on the amount of traffic that the company is getting to the bot. Right now we have a couple of paying clients, and now we're working on expanding and really growing the company and getting more clients on board. So by the end of our first year of actually selling the product, we're expecting to have about 7 clients who have paid us just about $800,000 in terms of revenue. And also at about that time, we're expecting to have $80,000 monthly recurring revenue. So here's our traction so far. So right now we're just about to close our second client and actually our second client is most likely gonna be a follow-on from our first client. So our first client is a really big company and they have a lot of different uses for our technology. So our second client is most likely gonna be in that same company because they really like what we've done and how we've built a bot for them. So what you're really looking at is a growth of our clients and our real end goal here is to get more monthly recurring revenue into the company to have a more stable amount of income. And the way that the company's been run so far has really worked out quite well for us because our CEO comes from a kind of a hardware background. So he really understands the complexities involved in actually making deep learning work on a hardware level. He used to work at NVIDIA. And our co-founder Dennis here, he is a CTO. And he has a PhD from the University of Toronto. So he's really experienced and he's been busy building the product and building all the algorithms from scratch. And that's something that a lot of companies don't do. They actually use other systems or like IBM or Google systems. We have our own in-house AI and NLP systems developed that we find actually more effective than some of the bigger company systems.

And that's it. Great, thank you very much. Boom. So, um, Ken asked, uh, there's a lot of questions already, but Ken asked, uh, is there an opening live demo where I can ask questions to test the product in real time? And if you want, you can paste it in the chat or just explain how to access it. How can I order beer, for example?

Oh, how can you order a beer? So that one hasn't gone live yet. What you can do is we usually have our bot on the site, but we're just making some upgrades to it right now. What you can do to test out some of the bots is you can go on Facebook and maybe you can look up Beer Bot. And I think there's a couple different Beer Bots, but eventually I think if I should, one of those should be the right one. So maybe I can post some links into the chat.

Yeah, post some links in the chat when we're done. Another kind of funny one from Andrew: Why is the company called DeepPixel if you're doing NLP? I don't see pixels in your data. Yeah, there's quite a few more, so why don't we do that one and I'll try and burn through this.

We get that question a lot actually, and it's kind of an inside joke. So deep comes from the deep learning aspect of what we do, and the pixel comes from— it's kind of a reference to GPUs and how they help with the AI and the deep learning space. So that's the kind of joke there, but yeah, it's not that funny.

So an inside joke. Okay. One of the kind of interesting one is, you know, how I want to know what deep spends, like, you know, how much work has gone into or how much money has gone into getting to where you are with this MVP was asking Josh.

So we spent about a year of time developing the core technology behind what we do. So I think the company really started with Yuri and Denis in their basement building out the real core of the technology. And just about in January of this year, we were ready to start actually going through some trials, and we went with a few different startups that we knew, and we started trialing the software. And now we've really gone to a mode where we're working on closing deals and generating revenue.

And Paul was asking, how are you different than something like other bot solutions with AI, maybe Watson, Bluemix, etc.?

Right, so we're actually in the process of working on a white paper that compares our technology to their technology. But I would say that one, with us you get more control, because if you're a company, and especially if you're a smaller company, if you want help from IBM or some of these other companies in terms of deploying the bot or fixing bugs in the system, They're gonna charge you an awful lot more than we will. And we also have a lot more control of the technology because we can change the specific instance of the bot tailored for you rather than with IBM. It costs you hundreds of thousands of dollars to get an instance running for yourself. So cost is a big thing for us. And so servicing a lot of these smaller needs. And we've actually seen our algorithm outperform some of the competitors. Because we actually use multiple different algorithms and we have a phrase-based approach to the AI where a lot of other companies actually just use a keyword-based AI approach. And so that's— those are the key differences, I would say. Okay.

There's a lot of questions. So I do want you to go in the chat when we're done. There's a couple that I like. Chris was asking, do you got— do you have enterprise sales experience on your team? And that relates to what you were just saying because you were talking about how the cost was low and David was also asking, so I'll bundle them, you know, $50K seems kind of steep. So are you really going after low-end customers with that price point? So maybe if you can talk about enterprise sales and the price point and that would be great.

Yeah. So that's a fair point. So yeah, right now we're going after much larger companies because we need the revenue to expand. But I think our really, our end goal is to really sell the solution downmarket in a year or two once we've had the learnings from working with these big enterprise clients. So yeah, the other question I think, sorry, can you repeat part of the question?

I think I missed the first part. So do you mean that, do you have enterprise sales skills on the team to go after these companies?

Right, right. So, you know, we obviously with our CEO, he's been in enterprise sales for a long time. And within NVIDIA, he actually worked on some projects between them and Google. So there's definitely a lot of enterprise sales experience. And yeah, there's— but we're always looking for more because it's tricky to find great salespeople. So if one of the people in the audience is really good at enterprise sales, we'd love for you to reach out.

Ah, well, there you go. That's a great point to leave it on at about