Daniel Bourke built an AI personal trainer bot in 15-20 hours using Chatfuel to help gym beginners. He wants 100,000 recurring users and believes messaging apps will dominate social media in five years. The bot sends workout reminders and personalized coaching through Facebook Messenger.
Chapters
Yo, what's going on everyone? Welcome back. I believe it is day 96 of The Daniel Bourke Show. Thank you so much for tuning in. Thank you so much for listening. What have I been up to today? Well, if you listened to yesterday's episode or watched yesterday's video, first of all, my car broke down while I was driving and recording the video. I shouldn't have been recording while I was driving, so don't take that as a good thing to do. I shouldn't have been doing that. But my car ran out of fuel because I was so distracted by doing a video. I really do need a self-driving car. So maybe that's something I could get a job in and start working towards. But nonetheless, over the last 2 days I've been a bit of a hermit. I've put my head down. I built a personal trainer Messenger bot. So what the hell is that? Well, yesterday I talked about Messenger bots, but today I finish it off. It's now a fully functioning working product, um, and it— I did it with a lot of help. You can go to chatfuel.com and just start building your own bot. It'll take about 7 minutes and you can build a really basic one. Mine's, mine's probably got 100 steps or so, um, 100 different things that you can do with it. Um, and it took me, I would say, probably about 20 hours total, maybe a bit less, probably about 15 hours. I did spend a lot of today on it, I'm only just wrapping up. I've gotta switch off tonight, otherwise I'll be up all night keep going on it. I'm gonna fix it up during the week and make it a bit better and implement new features and stuff like that. So hopefully can help some people work out. What's my goal with it? I wanna help 100,000 people work out with Messenger Bot. So far it's been out for about 2.5 hours and I've got a dozen people messaging it, getting some great feedback. When I really start to notice it will be when I start to get recurring users. I won't classify someone as messaging it once as helping them work out. If they've recurring messages, say 2 or 3 times with specific messages, then I'll count, like with a workout message, if they've continually done workout messages, then I'll count that as an active user. And so, That's my goal. I want to help 100,000 people work out, and I'm slowly working out how to do it. The way I figured, I've got to, I've got to build this for one individual person and then work out how to scale it. So right now, I would say my target customer for this, even though it's free, my target customer is someone who doesn't know too much about working out, and they might be just new to the gym. They don't know what their workout plan is. And they don't know the main exercises, so it can ask the bot the exercises. And within the next week or so, maybe, maybe 2 weeks, I don't know, probably less than a week actually, I'll try and implement some new features to send reminders of when someone should be working out or something like that. And I've got to work out how to do that first, but this is all a process. And by the way, what I'm doing is nothing, nothing special, by the way. It's Anyone can do it right now. You can jump on chatfuel.com. I'm not affiliated with them in any way, shape, or form, but I really think messenger bots and communication is, is where the next big boom is in tech, and that's with voice and text chat. I think text chat will be first, and then voice, once it sort of picks up— it's slowly getting there— that will, that will come out. That'll be really big in the next 5 years or so. Right now, I think the growth of the top 4 social media platforms, whichever they are, just got overtaken by the top 4 messaging apps. So I think that would be WhatsApp, Messenger, maybe WeChat and Line or something like that. So messaging apps are taking over social media apps. That's why I built a messenger bot, not like a website or something like that. So, or a fully blown app. Actually, I haven't built a fully blown app yet. It's just a, it's just a messenger bot and I did it within, within a couple of days. And I watched a guide on YouTube. So if you want to get this stuff like started, if you want to build your own, and I encourage you to if you want, like you can. It's just fun to have. I built in some random features into mine, like you can message it, "Yeah buddy," and it will send back an audio clip of Ronnie Coleman saying, "Yeah buddy." It's really fun, it's really fun, I gotta admit. And even going through the chat history of seeing what people actually say to it, and then I can I can see what people say to it, and then I go build that feature into the bot and then reply to them manually and say, hey, I just built this feature in, you can use it now, thank you for, for messaging the bot. So that's how I plan on building it up for the start, is just literally by hand going through individual messages and then building in the feature as, as they, uh, as they come. I might start signing off with my name so it's actually not just a bot always messaging, it's a human being. If I, if they, yeah, that's a good idea. Maybe add a personal touch to it. If they message the bot something it doesn't understand, and then I can reply back with it. But where can you find this bot? Well, I haven't got a customized Facebook page for it yet. It has a Facebook page, but the URL, I don't know what it is. I need to for some reason I need to have more likes or something like that, or maybe invite another admin to the page or something. I don't know. Facebook is weird. It's not letting me have a custom URL yet, but I will get there in the future. In the meanwhile, I've shared it on my social media, so facebook.com/MrDeBurke. M-R-D-B-O-U-R-K-E. I'll put a Messenger code on my website or something like that, MrDeBurke.com. I'll probably write a guide on Medium on how I did it within the next couple of days or something. But yeah, I'm really, I'm really liking it. I'm having, I'm having so much fun. I've got a lot of energy right now because it's— I think once you've built something of your own, you get this new, newfound energy. Like, all these features are running through my head. I've got like a little journal I'm writing down all the feedback I'm getting. I posted it in a big Facebook group of all the MIT Bootcamp alumni, and all of them are starting to comment on it and say, Hey, you could try this, you could try that, you could try that, you could try this. And I've got a lot of work to do. I've got a lot to work to. I just, I don't know, I set myself a challenge over the weekend and built a Messenger bot. Who knows? But it's still really early days. I've got a long way to go and I'm looking forward to it. What else is new? So, in the coming week I've got a lot of study to do, so I'm going to be splitting my time between study, Messenger bot, working out. I didn't even work out properly today. It was actually probably a good day to rest because my legs are really sore. So, resting up on the legs today. Working on my Udacity course tomorrow, deep learning. I'm going to the library with one of my friends. That should be a good time. I've got into a habit recently of going to the library to do study. And I take my laptop there without the charger because I'm using a MacBook Pro and I know the battery life's supposed to be all day, but what I do on it, some of the stuff is a bit programming intensive, like a bit CPU intensive, sorry. So the battery tends to drain in maybe 4 hours, 5 hours or so. I don't know if that's bad or whatever. I don't really mind. It's, I'm never really that far from a charger. And so that's my like limit to get out. So once my Mac sort of says, 'Your Mac is low battery, you got to get out.' That's been my window to sort of go home, have something to eat, and have a sleep. That's my tap-out schedule. So, I'll probably be spending a lot of time in the library over the next week. I got 4 more days left after this of— Day 100 is going to be this week. That's awesome. Yeah. Day 100 of the podcast experiment, but don't worry, I'm not stopping it there. I'm going to keep going, possibly not daily. No, actually not daily, weekly. I'll set you up for success now. I haven't worked out which day of the week yet. I want to get it a scheduled routine. As for YouTube videos, more of them are to come. I'm trying to work out this whole process of being creative, building products and whatnot. Loving it, but at the same time, you gotta dedicate a lot of time to it, but that's a good thing, right? I'm still learning. This is all the process. But whatever you're doing— oh, actually, what's the daily challenge? Hmm, try out my bot. Go to facebook.com/mrdburke and you can see the link, the latest link that I shared. You go to the Move More bot, give me some advice, give me some feedback. I'd really appreciate it. And yeah, with your help, we can make the world move more. Peace.
