What Computer Should You Buy for Local AI

I want to run AI locally. What computer I want to run AI locally. What computer should I buy? And that sounds simple should I buy? And that sounds simple should I buy? And that sounds simple enough. But it's not because the real enough. But it's not because the real enough. But it's not because the real answer depends on the model, the size, answer depends on the model, the size, answer depends on the model, the size, the quantization, how much VRAMm you the quantization, how much VRAMm you the quantization, how much VRAMm you have, whether you care about speed, have, whether you care about speed, have, whether you care about speed, whether you need a laptop, a Mac, or a whether you need a laptop, a Mac, or a whether you need a laptop, a Mac, or a PC. I've tested a ridiculous amount of PC. I've tested a ridiculous amount of PC. I've tested a ridiculous amount of hardware on this channel for this, and I hardware on this channel for this, and I hardware on this channel for this, and I feel like I'm just getting started. feel like I'm just getting started. feel like I'm just getting started. Macs, Nvidia GPUs, gaming laptops, mini Macs, Nvidia GPUs, gaming laptops, mini Macs, Nvidia GPUs, gaming laptops, mini PCs, workstation cards, weird little AI PCs, workstation cards, weird little AI PCs, workstation cards, weird little AI boxes, and giant desktop builds. But boxes, and giant desktop builds. But boxes, and giant desktop builds. But here is the problem. Most of the data is here is the problem. Most of the data is here is the problem. Most of the data is trapped inside my videos and maybe on my trapped inside my videos and maybe on my trapped inside my videos and maybe on my notes or somewhere on spreadsheets. So notes or somewhere on spreadsheets. So notes or somewhere on spreadsheets. So today I'm going to try and turn all that today I'm going to try and turn all that today I'm going to try and turn all that into an app. The idea is simple. You into an app. The idea is simple. You into an app. The idea is simple. You tell it what AI model you want to run, tell it what AI model you want to run, tell it what AI model you want to run, your budget, what kind of computer you your budget, what kind of computer you your budget, what kind of computer you want, and it tells you what hardware want, and it tells you what hardware want, and it tells you what hardware actually makes sense based on real actually makes sense based on real actually makes sense based on real benchmark data. And I'm building it in benchmark data. And I'm building it in benchmark data. And I'm building it in Replet. Replet is sponsoring this video, Replet. Replet is sponsoring this video, Replet. Replet is sponsoring this video, and they asked me to show what I could and they asked me to show what I could and they asked me to show what I could actually build with it. Now, I'm a actually build with it. Now, I'm a actually build with it. Now, I'm a software developer, but I still use AI software developer, but I still use AI software developer, but I still use AI to build a lot of my one-off to build a lot of my one-off to build a lot of my one-off applications, because you have to know applications, because you have to know applications, because you have to know where your time is most valuable, where your time is most valuable, where your time is most valuable, especially now when AI is just getting especially now when AI is just getting especially now when AI is just getting going. And I know that this is one of going. And I know that this is one of going. And I know that this is one of those things that I want to delegate, those things that I want to delegate, those things that I want to delegate, and I don't want to write myself from and I don't want to write myself from and I don't want to write myself from scratch, even though I enjoy coding, but scratch, even though I enjoy coding, but scratch, even though I enjoy coding, but this app can't just look nice. It this app can't just look nice. It this app can't just look nice. It actually has to get the recommendations actually has to get the recommendations actually has to get the recommendations right or close and then get tuned over right or close and then get tuned over right or close and then get tuned over time. So, let's get started. First thing time. So, let's get started. First thing time. So, let's get started. First thing I need to do is actually grab the data I need to do is actually grab the data I need to do is actually grab the data from some of my old videos. And I'm from some of my old videos. And I'm from some of my old videos. And I'm using AI to pull the data out. And it using AI to pull the data out. And it using AI to pull the data out. And it looks like this might be harder than I looks like this might be harder than I looks like this might be harder than I thought. I might actually have to watch

thought. I might actually have to watch thought. I might actually have to watch every one of these videos cuz in some of every one of these videos cuz in some of every one of these videos cuz in some of these I'm specifying the model, but these I'm specifying the model, but these I'm specifying the model, but maybe not the quantization. Bad, Alex. maybe not the quantization. Bad, Alex. maybe not the quantization. Bad, Alex. Or maybe the AI is not picking it up, so Or maybe the AI is not picking it up, so Or maybe the AI is not picking it up, so I have to scrub. But at least this gives I have to scrub. But at least this gives I have to scrub. But at least this gives me a starting point with some data me a starting point with some data me a starting point with some data unstructured. Even though it looks unstructured. Even though it looks unstructured. Even though it looks structured, it's pretty much structured, it's pretty much structured, it's pretty much unstructured. And we're going to feed unstructured. And we're going to feed unstructured. And we're going to feed this into Replet to see what we can this into Replet to see what we can this into Replet to see what we can start with. This is my workspace in start with. This is my workspace in start with. This is my workspace in Replet. And I'm really curious about the Replet. And I'm really curious about the Replet. And I'm really curious about the mobile app, 3D game animation. mobile app, 3D game animation. mobile app, 3D game animation. That would be really cool, but I have to That would be really cool, but I have to That would be really cool, but I have to I have to focus. It's sometimes hard to I have to focus. It's sometimes hard to I have to focus. It's sometimes hard to focus these days with all the different focus these days with all the different focus these days with all the different options I want to try. But website is options I want to try. But website is options I want to try. But website is what I'm making. It's a web app. I'm what I'm making. It's a web app. I'm what I'm making. It's a web app. I'm going to paste in a prompt that I going to paste in a prompt that I going to paste in a prompt that I actually generated. This is a trick that actually generated. This is a trick that actually generated. This is a trick that I use a lot. I don't write the prompt I use a lot. I don't write the prompt I use a lot. I don't write the prompt myself because I could forget some myself because I could forget some myself because I could forget some things. So what I do is I have the agent things. So what I do is I have the agent things. So what I do is I have the agent plan it out for me first and then based plan it out for me first and then based plan it out for me first and then based on the plan I have it generate an on the plan I have it generate an on the plan I have it generate an implementation prompt and that's what implementation prompt and that's what implementation prompt and that's what this is. Now there is a line in here this is. Now there is a line in here this is. Now there is a line in here that says do not invent benchmark that says do not invent benchmark that says do not invent benchmark numbers or fill missing results. Um I numbers or fill missing results. Um I numbers or fill missing results. Um I don't know about you but from my don't know about you but from my don't know about you but from my experience using AIs you're never going experience using AIs you're never going experience using AIs you're never going to trust it fully. It can create an to trust it fully. It can create an to trust it fully. It can create an incredible application for you, but the incredible application for you, but the incredible application for you, but the data the data that you're going to be data the data that you're going to be data the data that you're going to be surfacing and using that has to be surfacing and using that has to be surfacing and using that has to be validated. And even though I'm going to validated. And even though I'm going to validated. And even though I'm going to put that in there as part of my prompt, put that in there as part of my prompt, put that in there as part of my prompt, we're going to design with that in mind.

we're going to design with that in mind. we're going to design with that in mind. I don't really care so much about that I don't really care so much about that I don't really care so much about that because when I get my data, I'm going to because when I get my data, I'm going to because when I get my data, I'm going to validate it and then fill up the validate it and then fill up the validate it and then fill up the database with actual data. For now, it database with actual data. For now, it database with actual data. For now, it can use fake data if it wants to. Now, can use fake data if it wants to. Now, can use fake data if it wants to. Now, the first thing I'm going to do here is the first thing I'm going to do here is the first thing I'm going to do here is still I still wanted to plan because still I still wanted to plan because still I still wanted to plan because this agent has not had this prompt yet. this agent has not had this prompt yet. this agent has not had this prompt yet. So, it's a new context. I wanted to run So, it's a new context. I wanted to run So, it's a new context. I wanted to run through the plan first instead of doing through the plan first instead of doing through the plan first instead of doing the build step. And you can select which the build step. And you can select which the build step. And you can select which one you want and plan will ask questions one you want and plan will ask questions one you want and plan will ask questions and plan your work. Now, if I'm working and plan your work. Now, if I'm working and plan your work. Now, if I'm working with a designer, I can actually upload a with a designer, I can actually upload a with a designer, I can actually upload a design system here too. Things like design system here too. Things like design system here too. Things like colors, fonts, components, things that a colors, fonts, components, things that a colors, fonts, components, things that a designer would create in Figma, for designer would create in Figma, for designer would create in Figma, for example. But I don't have that. So, I'm example. But I don't have that. So, I'm example. But I don't have that. So, I'm going to skip that part. Let's go. Yes, going to skip that part. Let's go. Yes, going to skip that part. Let's go. Yes, we are starting it. Now, this agent is we are starting it. Now, this agent is we are starting it. Now, this agent is going to go through and it's going to going to go through and it's going to going to go through and it's going to take my data and try to find some gaps take my data and try to find some gaps take my data and try to find some gaps and inconsistencies, identify that and and inconsistencies, identify that and and inconsistencies, identify that and start working on this. Now, it suggested start working on this. Now, it suggested start working on this. Now, it suggested that this is a front-end only app that this is a front-end only app that this is a front-end only app because we're using all static data. because we're using all static data. because we're using all static data. That is something that you can decide on That is something that you can decide on That is something that you can decide on and steer the agent where you want it to and steer the agent where you want it to and steer the agent where you want it to go. Even though I've tested a lot of go. Even though I've tested a lot of go. Even though I've tested a lot of hardware here, I haven't tested that hardware here, I haven't tested that hardware here, I haven't tested that much hardware that it would require a much hardware that it would require a much hardware that it would require a database. all that work and it can database. all that work and it can database. all that work and it can probably fit just in one 5 kilobit JSON probably fit just in one 5 kilobit JSON probably fit just in one 5 kilobit JSON file, maybe even less. Looks like it's file, maybe even less. Looks like it's file, maybe even less. Looks like it's done. It found an appropriate domain for done. It found an appropriate domain for done. It found an appropriate domain for me that I can host it or I can do a me that I can host it or I can do a me that I can host it or I can do a custom domain up here if I want to under custom domain up here if I want to under custom domain up here if I want to under tools. So, there's preview. This is what tools. So, there's preview. This is what tools. So, there's preview. This is what it actually looks like. I might have to it actually looks like. I might have to it actually looks like. I might have to fix uh the light mode situation for some fix uh the light mode situation for some fix uh the light mode situation for some of you. That's triggering, I know. But of you. That's triggering, I know. But of you. That's triggering, I know. But under tools, we can have a custom domain under tools, we can have a custom domain under tools, we can have a custom domain here. There's publishing monitoring. So

here. There's publishing monitoring. So here. There's publishing monitoring. So you can view traffic and analytics. Now, you can view traffic and analytics. Now, you can view traffic and analytics. Now, it ended up actually using a database. it ended up actually using a database. it ended up actually using a database. So here is the databases section, and So here is the databases section, and So here is the databases section, and it's currently using 29 megabytes out of it's currently using 29 megabytes out of it's currently using 29 megabytes out of 20 gigs. I think uh I need to do a lot 20 gigs. I think uh I need to do a lot 20 gigs. I think uh I need to do a lot more benchmarking to fill up 20 gigs. more benchmarking to fill up 20 gigs. more benchmarking to fill up 20 gigs. Benchmarks, 15 rows, and it tried to do Benchmarks, 15 rows, and it tried to do Benchmarks, 15 rows, and it tried to do the best it could with the data that I the best it could with the data that I the best it could with the data that I gave it, the dirty, nasty data. It needs gave it, the dirty, nasty data. It needs gave it, the dirty, nasty data. It needs to be cleaned. Here's the hardware that to be cleaned. Here's the hardware that to be cleaned. Here's the hardware that it identified. It even left a space for it identified. It even left a space for it identified. It even left a space for memory because I did not actually have memory because I did not actually have memory because I did not actually have that in my data set, but I need to fill that in my data set, but I need to fill that in my data set, but I need to fill that in. So, DJX Spark has 128 GB. Save that in. So, DJX Spark has 128 GB. Save that in. So, DJX Spark has 128 GB. Save changes. So, I can update a database changes. So, I can update a database changes. So, I can update a database right there in line. M4 Pro has 64, 128 right there in line. M4 Pro has 64, 128 right there in line. M4 Pro has 64, 128 here, 32 for the 10 store. AMD 7900 has here, 32 for the 10 store. AMD 7900 has here, 32 for the 10 store. AMD 7900 has 32. And I'll have to figure out what 32. And I'll have to figure out what 32. And I'll have to figure out what that Nvidia GPU is referring to. Ah, the that Nvidia GPU is referring to. Ah, the that Nvidia GPU is referring to. Ah, the benchmarks table has the link to my benchmarks table has the link to my benchmarks table has the link to my YouTube video, which is perfect. And YouTube video, which is perfect. And YouTube video, which is perfect. And that's the 10 store video that I did. that's the 10 store video that I did. that's the 10 store video that I did. So, that GPU is the RTX 5090 32 gigs.

So, that GPU is the RTX 5090 32 gigs. So, that GPU is the RTX 5090 32 gigs. And actually, this is the AMD 9700. And actually, this is the AMD 9700. And actually, this is the AMD 9700. Let's save those changes and go back to Let's save those changes and go back to Let's save those changes and go back to preview. All right. Now, typically, preview. All right. Now, typically, preview. All right. Now, typically, right after you oneshot the application, right after you oneshot the application, right after you oneshot the application, you probably want to do a little review you probably want to do a little review you probably want to do a little review first to clean things up a little bit. first to clean things up a little bit. first to clean things up a little bit. So, I'm going to give the agent this So, I'm going to give the agent this So, I'm going to give the agent this task. Review the app as a skeptical task. Review the app as a skeptical task. Review the app as a skeptical local AI user. Identify five problems. local AI user. Identify five problems. local AI user. Identify five problems. Do not change the app yet. Just give me Do not change the app yet. Just give me Do not change the app yet. Just give me the plan and let's go. Now, I want you the plan and let's go. Now, I want you the plan and let's go. Now, I want you to notice that I've reached my monthly to notice that I've reached my monthly to notice that I've reached my monthly quota limit. This is what I got so far quota limit. This is what I got so far quota limit. This is what I got so far for it. So, if I want to keep working for it. So, if I want to keep working for it. So, if I want to keep working with it, I'm going to need to upgrade. with it, I'm going to need to upgrade. with it, I'm going to need to upgrade. That was a completely free plan, by the That was a completely free plan, by the That was a completely free plan, by the way. So, you can go down there, use the way. So, you can go down there, use the way. So, you can go down there, use the link down below, and subscribe and check link down below, and subscribe and check link down below, and subscribe and check it out. Run an app for yourself. But, it out. Run an app for yourself. But, it out. Run an app for yourself. But, I'm going to upgrade this for now. I'm I'm going to upgrade this for now. I'm I'm going to upgrade this for now. I'm going to go with Replet Core. And there going to go with Replet Core. And there going to go with Replet Core. And there I go. Welcome to Replet Core. This gives I go. Welcome to Replet Core. This gives I go. Welcome to Replet Core. This gives you workspaces so you can work as a you workspaces so you can work as a you workspaces so you can work as a team. But let's go to this project that team. But let's go to this project that team. But let's go to this project that I was doing. By the way, it was not all I was doing. By the way, it was not all I was doing. By the way, it was not all that for the month. I also created this that for the month. I also created this that for the month. I also created this app to kind of see what it's capable of app to kind of see what it's capable of app to kind of see what it's capable of before doing the one I showed you. And before doing the one I showed you. And before doing the one I showed you. And this is like a little YouTube clone that this is like a little YouTube clone that this is like a little YouTube clone that I did. Not bad. But let's get back to I did. Not bad. But let's get back to I did. Not bad. But let's get back to our task. Continues on and it gave me a our task. Continues on and it gave me a our task. Continues on and it gave me a prioritize problem list. The core prioritize problem list. The core prioritize problem list. The core feature is broken. So, let's do that.

feature is broken. So, let's do that. feature is broken. So, let's do that. Address the first issue and actually Address the first issue and actually Address the first issue and actually address the second issue as well. Boom. address the second issue as well. Boom. address the second issue as well. Boom. I want to show you that because Replet I want to show you that because Replet I want to show you that because Replet can actually work on two things at once can actually work on two things at once can actually work on two things at once in parallel because it can split up into in parallel because it can split up into in parallel because it can split up into multiple sub agents and that's what it's multiple sub agents and that's what it's multiple sub agents and that's what it's doing right here. I can work on both in doing right here. I can work on both in doing right here. I can work on both in parallel. Let me read the current spec parallel. Let me read the current spec parallel. Let me read the current spec schema while prepping the frontend fix. schema while prepping the frontend fix. schema while prepping the frontend fix. Cool. I had to go through and fix all Cool. I had to go through and fix all Cool. I had to go through and fix all the other issues and for now this seems the other issues and for now this seems the other issues and for now this seems to work. So we select the model here. to work. So we select the model here. to work. So we select the model here. Everything else is optional. I'm just Everything else is optional. I'm just Everything else is optional. I'm just going to try this out. Get going to try this out. Get going to try this out. Get recommendations and yeah there it is. recommendations and yeah there it is. recommendations and yeah there it is. Tenstor Wormhole N300. It's a dev board Tenstor Wormhole N300. It's a dev board Tenstor Wormhole N300. It's a dev board recommended. Memory should be listed recommended. Memory should be listed recommended. Memory should be listed there, but it's not. But that's okay. there, but it's not. But that's okay. there, but it's not. But that's okay. We'll fix that. This is an iterative We'll fix that. This is an iterative We'll fix that. This is an iterative approach just like actual software approach just like actual software approach just like actual software development. But to get something like development. But to get something like development. But to get something like this, if I had to hand code this would this, if I had to hand code this would this, if I had to hand code this would probably take me hours. Literally like probably take me hours. Literally like probably take me hours. Literally like even though I have 25 years of software even though I have 25 years of software even though I have 25 years of software dev experience, this would have taken me dev experience, this would have taken me dev experience, this would have taken me some time. And and I'll be honest with some time. And and I'll be honest with some time. And and I'll be honest with you, I love you all and I want you to you, I love you all and I want you to you, I love you all and I want you to have these results, but I'm not going to have these results, but I'm not going to have these results, but I'm not going to sit there for many, many hours coding sit there for many, many hours coding sit there for many, many hours coding this thing up, even though I really want this thing up, even though I really want this thing up, even though I really want to. That's why I'm doing this. That's to. That's why I'm doing this. That's to. That's why I'm doing this. That's why I'm using an agent to do it for me.

why I'm using an agent to do it for me. why I'm using an agent to do it for me. Now, I want to do dark mode cuz uh you Now, I want to do dark mode cuz uh you Now, I want to do dark mode cuz uh you know, you all asked me about that. Add know, you all asked me about that. Add know, you all asked me about that. Add dark mode, light mode switch. Make dark dark mode, light mode switch. Make dark dark mode, light mode switch. Make dark mode the default. Boom. Worked for 1 mode the default. Boom. Worked for 1 mode the default. Boom. Worked for 1 minute. And there it is. Light mode, minute. And there it is. Light mode, minute. And there it is. Light mode, dark mode switch. Sorry about that. dark mode switch. Sorry about that. dark mode switch. Sorry about that. We'll keep it on dark mode for now. 1 We'll keep it on dark mode for now. 1 We'll keep it on dark mode for now. 1 minute. Now, one thing that Replet does minute. Now, one thing that Replet does minute. Now, one thing that Replet does that some other providers don't do at that some other providers don't do at that some other providers don't do at all is app testing. This is where the all is app testing. This is where the all is app testing. This is where the agent can be told to go and test the app agent can be told to go and test the app agent can be told to go and test the app out itself using an actual browser. out itself using an actual browser. out itself using an actual browser. Think of it like a stress test. So, I'm Think of it like a stress test. So, I'm Think of it like a stress test. So, I'm going to give this prompt right here. going to give this prompt right here. going to give this prompt right here. Test the entire recommendation flow Test the entire recommendation flow Test the entire recommendation flow using realistic and intentionally using realistic and intentionally using realistic and intentionally difficult user scenarios. Check the form difficult user scenarios. Check the form difficult user scenarios. Check the form results comparison view. Missing data results comparison view. Missing data results comparison view. Missing data behavior. We're going to have some of behavior. We're going to have some of behavior. We're going to have some of that, I have a feeling. Also, mobile that, I have a feeling. Also, mobile that, I have a feeling. Also, mobile layout. That's something you should care layout. That's something you should care layout. That's something you should care about if you're doing a web app. Is this about if you're doing a web app. Is this about if you're doing a web app. Is this web app responsive? And the answer is web app responsive? And the answer is web app responsive? And the answer is yes. Yes, it is, but somewhat limited. yes. Yes, it is, but somewhat limited. yes. Yes, it is, but somewhat limited. So, let's see if this test identifies So, let's see if this test identifies So, let's see if this test identifies that issue. It actually opens up a that issue. It actually opens up a that issue. It actually opens up a browser window right here in the browser window right here in the browser window right here in the sidebar, and you can see it clicking on sidebar, and you can see it clicking on sidebar, and you can see it clicking on different parts of the app. Is that different parts of the app. Is that different parts of the app. Is that Chrome? It's like a mini little robot Chrome? It's like a mini little robot Chrome? It's like a mini little robot using my app. It's cute. Okay, 7 minutes using my app. It's cute. Okay, 7 minutes using my app. It's cute. Okay, 7 minutes later, and it's done all the tests. 11 later, and it's done all the tests. 11 later, and it's done all the tests. 11 scenarios passed cleanly. And check this scenarios passed cleanly. And check this scenarios passed cleanly. And check this out. Yes, look at that. That responsive out. Yes, look at that. That responsive out. Yes, look at that. That responsive navigation menu at the top. That is navigation menu at the top. That is navigation menu at the top. That is beautiful. The admin panel, something I beautiful. The admin panel, something I beautiful. The admin panel, something I haven't showed you yet. This allows me haven't showed you yet. This allows me haven't showed you yet. This allows me to actually import real data and it gets to actually import real data and it gets to actually import real data and it gets processed and saved to the database. And processed and saved to the database. And processed and saved to the database. And um anybody can just load their own

um anybody can just load their own um anybody can just load their own stuff. I'll have to secure it later. stuff. I'll have to secure it later. stuff. I'll have to secure it later. Okay. Okay. Okay. So, this is the recommendation engine. So, this is the recommendation engine. So, this is the recommendation engine. This is the hardware that was used and This is the hardware that was used and This is the hardware that was used and these are the benchmarks. I think it's a these are the benchmarks. I think it's a these are the benchmarks. I think it's a very good start and I'm going to publish very good start and I'm going to publish very good start and I'm going to publish this so that you can all check it out. this so that you can all check it out. this so that you can all check it out. This does need tweaking and I'm going to This does need tweaking and I'm going to This does need tweaking and I'm going to continue using this agent to make those continue using this agent to make those continue using this agent to make those adjustments. Now I'm going to enable adjustments. Now I'm going to enable adjustments. Now I'm going to enable this feedback widget so that you can this feedback widget so that you can this feedback widget so that you can also submit your feedback here things also submit your feedback here things also submit your feedback here things you want to see. Now while it's doing you want to see. Now while it's doing you want to see. Now while it's doing some publishing you can check out logs. some publishing you can check out logs. some publishing you can check out logs. You can manage the virtual You can manage the virtual You can manage the virtual infrastructure that's running your infrastructure that's running your infrastructure that's running your application. You can see that all this application. You can see that all this application. You can see that all this so far cost me $341. so far cost me $341. so far cost me $341. Congratulations, your project is live. Congratulations, your project is live. Congratulations, your project is live. Let's go check it out. And there it is. Let's go check it out. And there it is. Let's go check it out. And there it is. Let's just do a quick check here. Llama Let's just do a quick check here. Llama Let's just do a quick check here. Llama 3.1. Get recommendations. And there's my 3.1. Get recommendations. And there's my 3.1. Get recommendations. And there's my recommendation. Let's select something recommendation. Let's select something recommendation. Let's select something else. Llama 3.23B. else. Llama 3.23B. else. Llama 3.23B. And yeah, pretty much everything can run And yeah, pretty much everything can run And yeah, pretty much everything can run that. Whoa, there. Okay, calm down, that. Whoa, there. Okay, calm down, that. Whoa, there. Okay, calm down, Alex. Don't blind everybody. That toggle Alex. Don't blind everybody. That toggle Alex. Don't blind everybody. That toggle switch works. hardware benchmarks and switch works. hardware benchmarks and switch works. hardware benchmarks and let's do F12 developer tools. Click this let's do F12 developer tools. Click this let's do F12 developer tools. Click this little button to take a look at what little button to take a look at what little button to take a look at what this looks like on a I don't know iPhone this looks like on a I don't know iPhone this looks like on a I don't know iPhone 14 Pro Max. Nice. The responsiveness 14 Pro Max. Nice. The responsiveness 14 Pro Max. Nice. The responsiveness works beautifully. I got a lot done in works beautifully. I got a lot done in works beautifully. I got a lot done in the last half an hour, 40 minutes. Now, the last half an hour, 40 minutes. Now, the last half an hour, 40 minutes. Now, there's more work to be done here, but there's more work to be done here, but there's more work to be done here, but that's how Replet turns an idea into an that's how Replet turns an idea into an that's how Replet turns an idea into an actual product that you can use. Use my actual product that you can use. Use my actual product that you can use. Use my link down below to try out Replet. You link down below to try out Replet. You link down below to try out Replet. You can start out for free. And I'll also can start out for free. And I'll also can start out for free. And I'll also link to this app so you can go play

link to this app so you can go play link to this app so you can go play around with it too. Otherwise, thanks around with it too. Otherwise, thanks around with it too. Otherwise, thanks for watching the video. You might also for watching the video. You might also for watching the video. You might also enjoy this one. And I'll see you next enjoy this one. And I'll see you next enjoy this one. And I'll see you next time.