I Built a Docusign Clone with Lovable.ai Because of Greg Isenberg | Vibe Coding

Description

This video is a "vibe coding" session in which I built a DocuSign clone inspired by a Greg Isenberg post I saw on LinkedIn. If you don't get bored easily and want to see if I was able to build this using only Lovable, this 40-minute session is a delight.

💬 Have any questions? Drop them in the comments! If this helped, don’t forget to like & subscribe for more app-building insights.

My links
LinkedIn: https://www.linkedin.com/in/guil-meira/
Instagram: https://www.instagram.com/guimeira/
Website: https://techmakers.dev/
ScopesFlow: https://scopesflow.com/

Summary

Building a DocuSign Clone with Lovable.AI in Under 40 Minutes

In this engaging "vibe coding" session, the creator demonstrates how to build a functional DocuSign clone using Lovable.AI, inspired by Greg Isenberg's challenge to create a free alternative to the expensive digital signature platform.

The developer begins by using ChatGPT to generate a comprehensive description of what DocuSign does, strategically renaming it "SignFlow." This description is then fed into the creator's own tool called ScopesFlow, which generates a detailed project scope including pages, features, APIs, and required assets. The resulting prompt is transferred to Lovable.AI to automatically build the application.

Despite encountering several technical challenges with browser performance and error handling, the session showcases how AI can rapidly develop a working prototype. Within 35 minutes, the creator successfully builds a digital signature platform featuring:

- A professional landing page with modern, Apple-inspired design
- User authentication system
- Document dashboard with filtering capabilities
- Document signing interface with signature placement
- Document management features (upload, send, organize)

The video demonstrates the impressive capabilities of AI-powered development tools, as the creator notes that what would have previously taken a week of traditional development was accomplished in just over half an hour with a single prompt. While the application isn't production-ready and requires further refinement, including database integration with Supabase and additional security features, the foundation is remarkably solid.

The creator explains how this workflow typically continues by sending the Lovable.AI-generated frontend to Cursor (another AI tool) for backend development and optimization. This approach represents a significant evolution in software development, making complex applications accessible to a wider range of creators.

For viewers interested in AI-powered development, document management solutions, or building SaaS applications, this video provides valuable insights into how modern tools can dramatically accelerate the development process.

Transcript

0:00 so I saw Greg Eisenberg um post a couple 0:04 weeks ago um where he uh was basically 0:08 saying that we should um create um a 0:13 clone for duus 0:15 sign um a free version a free 0:18 alternative because we are paying 0:21 everyone's paying a lot of money to uh 0:24 to use D sign but now with AI Revolution 0:28 we should just build our own for free 0:31 and then I saw his video using uh menus 0:35 AI where they tried to build but it did 0:40 a b a bunch of research checked all uh 0:44 documentation about doc sign but at the 0:46 end it wasn't able to build the app so I 0:50 thought okay I just build a tool where 0:53 we can basically create a prompt based 0:57 on um on uh scope of a project and what 1:02 if I ask chpt to build me a description 1:07 for do for duu sign clone and then add 1:10 to my tool and with my tool send to 1:15 lovable and then we would have like a 1:17 better prompt with every feature every 1:21 every page and everything else that we 1:22 need to build a dock Cyclone so this is 1:26 what we are going to do today first step 1:29 is is basically ask 1:32 jpd to 1:34 describe uh what DocuSign 1:37 is um I need you to give me 1:45 a 1:47 complete 1:49 description of 1:52 what docu sign 1:57 is this 2:01 should this should be only maybe three 2:09 sentences 2:11 and please do not 2:15 use doc docu sign name in the 2:23 description 2:25 create 2:26 your own name 2:31 okay let's see what jpt will give 2:33 us share signf flow is a secure digital 2:37 signature platform that allows 2:39 individuals and businesses to electronic 2:42 sign send and manage documents from 2:45 anywhere it string line workflows by 2:47 automating contract approvals 2:49 integrating with um popular tools like 2:52 CRM and cloud storage services and 2:55 ensuring legal compliance with 2:57 encryption and audit trails 3:00 with features like book sending mobile 3:04 accessibility and identity verification 3:07 it 3:08 enhance efficient efficiency while 3:11 reducing paperwork and turnaround times 3:15 okay this is pretty pretty good so we go 3:19 to my tool 3:21 here and we create a new project so let 3:25 me add the description and the name of 3:28 the project will be 3:30 sign flow for some 3:33 reason the name of the the my project is 3:36 Scopes flow and chat gave me sign flow 3:40 maybe it's just coincidence or maybe 3:42 it's the universe now thei will 3:45 generate the scope for us the flowchart 3:49 with the scope for us we will give just 3:52 a brief um uh quick look and I'm now 3:57 going to change add pages add features 4:01 change apis we probably will see some 4:03 crazy apis that I never um heard about 4:07 but let's see here open project okay 4:09 that's not 4:12 like that 4:13 big 4:15 um 4:17 okay so let's see first what do we have 4:21 here so we have a homepage we have a 4:23 user dashboard we have a document 4:25 signing we have a book send we have 4:28 profile settings we have user 4:31 authenication document management we 4:34 have document sign we have book document 4:38 sign we have docu sign API send GD API 4:43 AWS 4:45 API and um let's 4:50 see okay I'm not sure that I want to use 4:53 doign 4:55 API 4:56 because we want to build our own so I 4:59 will delete 5:01 this okay we're not going to use this 5:05 one 5:06 and because this is only a test what 5:09 what what what I would probably I would 5:12 probably do if what's something that I I 5:16 wanted to create like a real side 5:19 project I would just add more pages 5:23 probably and iterate more with the tool 5:27 to get like we don't have we don't want 5:29 use um um docu sign API of course so how 5:34 we're going to manage this we probably 5:36 have some errors 5:38 here but let's see now we have here the 5:42 um the description of the app this 5:44 description was created by uh uh by The 5:48 Prompt this is not our uh description 5:51 then we have here all the features we 5:53 have the pages we have the apis that we 5:56 want to use and we have the assets that 5:59 we need for for this project I like to 6:01 let the um um I wanted to add the assets 6:05 here because then the project knows that 6:09 we need a place for the app logo we need 6:11 to find a libraries for the icons and we 6:14 need graphics for the um for the landing 6:18 page so it will um uh give us places to 6:22 place this on the landing page so I'll 6:25 just cop 6:26 this go to lovable 6:30 paste 6:31 here and hit enter now we can wait and 6:35 see the magic 6:37 happen let's see if we can actually 6:40 build H doc sign clone before Greg 6:44 Eisenberg does um 6:49 okay this is just a blank page 6:55 uh okay we probably had an error 7:01 here let's try 7:04 to 7:05 okay come on this is the first 7:08 error let's try to copy and paste 7:23 again 7:25 okay let me see I have 7:30 credits let's refresh the 7:35 page okay maybe just maybe it's just too 7:39 much for l b when I tried the first time 7:43 my two it worked so let's see if we can 7:50 actually okay it's not even letting me 7:53 uh hit enter here 8:00 let's go back to the dashboard and let's 8:02 try 8:05 again okay so it looks like it's 8:08 probably some error with 8:12 um with 8:16 lovable let's try one more time let's 8:19 give one more shot 8:21 here let's now click here 8:27 again okay now it looks like that is 8:32 thinking spin Up 8:35 preview we see here this dot here that 8:38 does it look like is okay now now now we 8:42 have something okay we have the 8:44 structure okay we have the structure for 8:46 the landing page we have the structure 8:48 for the dashboard we have our index page 8:52 we have dashboard page we have assigned 8:54 document page we have off page it's good 9:00 um 9:04 okay we had the layout let let's see 9:10 if okay has always is using 9:15 Tailwind 9:16 [Music] 9:18 um and I do prefer that it us stay wind 9:21 because when everyone was using um vzero 9:26 I was using vzero so I was creating the 9:29 layout the design in visero and then 9:32 sending to cursor the problem was that 9:35 every project looked the same because it 9:37 was using Chetan 9:40 mostly so like it doesn't matter if I 9:43 was building um uh CRM or if I was 9:47 building um I don't know a Marketplace 9:51 or a social media they all look the same 9:55 so it's it's it's cool that it now is um 10:00 it's trying to build something that is 10:02 not like this look like everyone is 10:05 building the same 10:06 thing so okay 10:09 sticking um stick more time than I 10:12 expected in something like the folder I 10:15 don't know why let me see if I can go to 10:21 the okay now went to 10:24 the hero uh component for the homepage 10:38 okay I will let the video running 10:41 because I want then to um be able to 10:44 check how many um the minutes that we 10:48 spent here let me 10:57 just okay um Google Chrome was already 11:00 taking too much uh from my 11:04 laptop and now with loveable when I open 11:08 lovable in two tabs and I 11:11 open um cursor it's just like looks like 11:15 I can fry a EG here but okay it's going 11:20 it's 11:21 going it's spending more time than I 11:23 expected again and the the component for 11:26 the landing page but probably because on 11:30 our prompt we have here let me show 11:36 you we have 11:39 here the 11:42 homepage we have all these sections we 11:45 and we basically have two pages so yeah 11:49 it it's it's it's it's reasonable that 11:51 we will spend more time trying to um 11:56 improve the only pages that we have so 11:58 okay that's that's fine that's 12:01 fine let's see what we will get from 12:07 this 12:12 okay let me try to go down 12:16 again let's go let's go let's 12:19 go okay it's already on working on the 12:24 off I never know if I'm pronouncing 12:27 correctly sorry English is not my 12:30 primary 12:32 language and lately it's not even my 12:36 secondary language Spanish is 12:39 so yeah we are already in here so if is 12:45 following this um this 12:49 order we are already creating the um 12:52 components for the off page 12:57 so we should be almost at the end let's 13:04 see and I saw I think the difference uh 13:08 there something that would be pretty 13:09 cool if we have if we had this in um in 13:15 um in lovable is or even though this is 13:19 maybe something that I can add into my 13:22 app 13:24 because what I saw here from man AI is 13:27 um that it spent a lot of time doing 13:31 research and like taking screenshots um 13:36 uh scraping the page to find information 13:39 about the cosign checking probably 13:41 checking the the layout checking like we 13:43 see here the features checking the um 13:47 the whole structure that it used 13:50 probably it it wasn't like that um 13:53 accurate 13:54 because it wasn't able to log in into 13:57 DocuSign to see all the pages 13:59 and how worked the the UI and ux inside 14:03 the doshine dashboard but if we could 14:06 have um here something that is um a 14:11 search before starting building the app 14:14 okay we have we have an 14:17 error let's dismiss this is probably 14:23 something because of 14:25 my Google 14:27 Chrome so signing 14:30 [Music] 14:31 in 14:33 okay okay it looks like that we have um 14:37 he created an off 14:39 component and now is creating a sign in 14:44 component this is probably for the 14:46 signature not the um 14:49 authentication but let's see okay 14:54 this uh I hope this wasn't too much to 14:57 ask from uh from lovable let me close 15:00 these tabs here cuz I also want to 15:04 show uh the test that I 15:07 did that you can see in my uh my 15:15 YouTube 15:17 um let me 15:20 just because the first test that I did 15:23 with my to and um 15:26 lovable was asking to create like um a 15:30 Jito 15:32 album and it went pretty well like 15:36 better than what um I expected like way 15:41 better than what I 15:42 expected 15:44 um like it created the um I was thinking 15:48 about building this in FLIR flow like 15:51 the project that I'm I'm talking about 15:54 um because I wanted to be um mobile 15:57 first I want to be a native app but I 16:00 even thinking now if I can build this in 16:05 lovable if is reactive and 16:10 um then send to cursor and ask 16:15 to to uh refractor and use 16:21 Expo um but I'm I'm still thinking about 16:24 this let this let wait this for uh load 16:28 here 16:32 okay now it went back to the dashboard 16:36 page and here we are still loading I 16:40 probably need an uh a new 16:43 laptop or maybe just 16:46 buy uh standard computer with a better 16:52 um with a better setup let's 16:57 see come on on I really want this to 17:01 work cuz here man it looks so cool to be 17:04 honest not that is not because it's my 17:08 tool not because I built this 17:13 but this is something that I saw someone 17:17 uh on um Twitter saying that this would 17:20 be pretty cool to have on um on 17:25 loveable like a place where we could 17:28 create Pages create the features be able 17:31 to see the S map before start buildings 17:34 before start 17:35 prompting but 17:38 okay um yeah like I 17:42 said it was to 17:45 much um a client expection see the 17:49 browser console for more information 17:52 okay 17:53 probably because I had to this two tabs 17:57 open 17:59 let's 18:03 um okay created okay this one was 18:07 running to 18:09 probably let's see the console give me 18:12 five minutes and I will figure this 18:16 out 18:26 okay okay let's just close close this 18:29 one close this one I will not give up 18:33 for real I will not give up I will make 18:36 this work I will make this 18:41 work okay now looks like I can see my 18:44 project okay it's just trying to 18:47 load come on come 18:50 on so this is the one that I was 18:52 building right let's click the 18:57 here uh 19:00 um okay it looks like they created the 19:04 dashboard now created the sign in 19:10 now okay um build wasn't 19:13 successful cannot read package let's 19:16 just try to fix 19:19 this what let's see what would 19:25 happen cuz nothing is actually working 19:30 and and here it says I have created a 19:33 stunning digital signature 19:36 platform with a modern clean interface 19:39 inspired by Apple's design now I really 19:43 want to see 19:45 this now now now we we need to see 19:50 this I I we would had to find a way to 19:53 fix this for 19:57 real not not even if I need to go 20:00 somewhere and buy a computer we are 20:04 going to finish 20:06 this let's see if we can do this without 20:08 me 20:10 spending tons of 20:12 money to buy a computer that can then 20:17 handle this let's 20:22 see okay it went back to the index 20:26 page so this this is something that 20:31 um uh it's probably yeah it's probably 20:35 something wrong with the client side so 20:37 probably something wrong with my um 20:42 laptop it's not being able to handle 20:45 this but it would be cool if anyone 20:51 could do this so anyone with any laptop 20:55 any computer any like setup set up could 20:59 do this so then we could do this using 21:02 our mobile 21:04 devices and work on our mobile devices 21:07 it's something that I don't like 21:10 not I don't really want to work on my 21:14 phone like build an app on my phone but 21:20 maybe if I have an idea and I want to 21:23 implement right away this could probably 21:26 something uh be something that we could 21:28 do 21:31 but let's uh see if now it's going to 21:35 work because it started creating like 21:38 like we said before we had index page we 21:41 have dashboard we have um the sign 21:43 document page we see here that we 21:45 already have the uh the path so we have 21:48 the sign slash the ID of the document 21:52 that we are signing and we have the off 21:54 page 21:59 uh okay went back to the um to the off 22:05 page I just wanted to like to be sure 22:09 that it just stop it working because of 22:14 my 22:15 [Music] 22:17 um my laptop and not because it was like 22:22 Nar with the 22:24 project I have moved all the page 22:26 components into 22:28 okay this is 22:32 reasonable I 22:34 added the mission file 22:37 text let's see the code where is 22:42 my okay the build was sucessful 22:46 again uh we have the call in here 22:51 so or testimonials are all inside 22:56 components home 22:59 component home we have 23:01 here features we have hero and we 23:08 have testimonials okay let's see the 23:12 error that we are 23:15 getting okay 23:17 because probably if we see here that's 23:20 something that I learned um not a long 23:23 time ago but if we come here 23:32 uh actually let's just this is this 23:37 video is already too long oh 23:42 okay 100,000 23:49 characters 23:53 okay um 23:59 let's try first to refector the 24:04 dashboard maybe if we break down the 24:06 dashboard we will be able 24:09 to um to fix 24:15 this bringing into small components okay 24:19 this this is something that 24:22 um this is is also a problem with cursor 24:26 I'm not sure if the other um 24:28 platforms if like boat or repet do the 24:33 same but usually creates huge files like 24:38 huge files um I Tred to build 24:42 something um in the cursor and I 24:47 remember like the uh again the dashboard 24:52 dashboard is Al always the problem but 24:54 the dashboard had like 600 700 lines of 24:59 gold and we had a lot of things that 25:02 could be a component being called and I 25:06 had to then I I had to ask to refactor 25:10 the dashboard didn't give me cursor 25:13 didn't give you um cursor don't give you 25:16 like these cool buttons here saying hey 25:19 maybe you should refactor the dashboard 25:21 or refactor the 25:23 index but let's see if it will be able 25:26 to refractor the de dashboard at 25:30 least um but but from what I can see 25:36 here probably so if we cannot find 25:41 module cannot find the 25:43 module 25:51 um okay we have a different error now 25:58 oops looks like our AI had a bit of hit 26:00 up hit 26:09 cup okay it looks like this is not going 26:13 to work for some 26:20 reason I believe that this was too much 26:24 for the 26:26 AI we're not going to be able to fix 26:30 this this is the last last 26:33 try this is the last try okay may maybe 26:38 maybe Greg Eisenberg gave us this task 26:41 because he knew this was too much he 26:44 knew that this was something that we 26:46 could not do with just one 26:50 prompt let's see 27:08 um what I'm not understanding 27:13 is we 27:15 have for for our Pages we have off we 27:20 have dashboard we have index we have um 27:23 not found and we have um signed up 27:28 document for sign document we have 1 2 3 27:33 4 5 27:38 six 27:39 seven Imports so not that many and one 27:45 is for uh an effect one is for uh 27:49 navigation one is for animation one is 27:53 for um uh the icons one is for a button 27:58 and just one is an actual 28:02 component 28:10 so 28:18 okay if maybe if we take this and send 28:24 this the way that it is now and send to 28:29 cursor cursor would probably have a 28:34 better um perspective on how to manage 28:39 this cuz now I think this this is not uh 28:42 a front end 28:44 problem this is 28:46 probably um something that not like for 28:50 me lovable is my um front end developer 28:55 and cursor is my backend developer 28:58 anytime that I want to improve um 29:01 security improve performance I take from 29:05 lavable and send to cursor but usually I 29:10 do this after the features are set up so 29:15 after I can at least um do some 29:17 navigation I can see the features how 29:19 the features work and we have the design 29:23 done because cursor even though I always 29:26 ask to keep uh using the same design 29:30 usually just create crap design for some 29:34 reason maybe I'm I'm 29:36 not prompting correctly but let's 29:54 see okay now it's telling me that 29:58 uh I fixed all the Errors By remove the 30:03 duplicated file 30:05 text created all the missing module 30:09 files 30:12 um yeah I don't think that 30:16 okay okay created the Ling page that's 30:19 good okay it's a pretty cool landing 30:25 page not going to lie okay 30:30 but what we want to know is if we can 30:34 actually use this I'm not going to be 30:36 able to sign in because it's not 30:37 connected to any of 30:40 my 30:44 um uh super base 30:50 projects and yeah the 30:54 dashboard okay now the dashboard is 30:56 broke dashboard is broke let's 31:00 see now the error is way way shorter 31:04 than it was before uh let's see if we 31:07 can 31:09 um if we can fix this now 31:14 um I just want to say sorry for 31:19 um um to lovable team CU it looks like 31:25 that I'm forcing the AI to much much I 31:29 will probably break this at some point 31:32 I'm and I'm I'm sorry uh let me check 31:35 the file to fix and the issue is fixed 31:39 by changing the default value of 31:45 terms um not sure if I 31:50 follow the off form okay 31:53 so probably is not showing because 31:58 we are not actually signed in right 32:03 because the schema defines terms has 32:06 little true means that it can only be 32:10 true by setting the default value true 32:13 you can 32:16 true 32:18 okay can we fix the this is the first 32:24 time that I'm typing here so after I I 32:26 copy and paste here 32:28 the The Prompt the only thing that I had 32:31 to do was just click to try to fix the 32:35 error can we fix the 32:38 dashboard so we can take a 32:43 look please 32:46 remember this project is not yet 32:52 connected to any 32:56 database we we 32:59 will connect to super 33:03 base 33:05 once we can at 33:10 least 33:12 navigate the 33:15 dashboard oops 33:21 dashboard now let's see if can fix the 33:24 dashboard because 33:29 two two two two options here one we are 33:32 not being able to see the dashboard 33:34 because we are not locked in but we 33:36 can't log in like the the app can't 33:39 verify that we are logged in because we 33:43 have the um the off form but we don't 33:47 have connected to to to super base so we 33:50 can 33:51 actually um authenticate the 33:54 user maybe this is the problem or 33:58 maybe it fixed the index page it fixed 34:01 the off page but it didn't fix the 34:03 dashboard page because if we take a look 34:06 in 34:08 here uh the one that is working is 34:11 index so we have all the um components 34:16 here okay we have all the components 34:19 here imported and but then here in the 34:23 dashboard we have navbar we have header 34:27 we have we have filters we have 34:29 documents grid we have quick actions we 34:32 have um 34:34 upload to upload the document and then 34:37 we have the actual document 34:40 card but I'm not sure if we have all 34:43 these 34:46 components 34:48 uh where is the okay we have the 34:51 document card let's see if fix it okay 34:54 okay okay never mind okay fix it 34:58 okay fix 35:00 it but if I try to ve here we are not 35:04 be okay we are able to see 35:08 this it's not is not 35:12 [Music] 35:15 um the best but 35:18 oh can we see the details 35:21 here now I can see the 35:25 signature if I click to 35:32 F 35:34 okay so this is like if I if I'm if I'm 35:37 logged in I just click here for the 35:40 signature just click here to add the DAT 35:42 and this just click here because I 35:44 already know my full name what if I 35:47 click download nothing 35:50 happens but we have here the good fing 35:54 we have documents we have all documents 35:56 we have drafts we have pening we we have 35:57 complete we can upload uh 36:01 document we actually can uh let's upload 36:05 like something that I don't know what it 36:07 is it's 36:08 uploading I'm not going to sure if it's 36:10 going to upload because we don't have a 36:12 database 36:15 so 36:17 okay kind of 36:20 did send for Signature books and 36:24 templates new document we have here our 36:27 profile so I believe that like we spend 36:31 now 35 36:34 minutes of course we have a lot to do 36:37 like a lot to do but if we imagine that 36:42 um a couple months ago with 36:46 um just this in um with no code with 36:51 bubble let's say um we will probably 36:56 spend like 37:00 I would spend a week building just this 37:04 let's see okay a weake working from 9 to 37:08 5: and not worrying too much about it it 37:13 will take me a week but now we spend 37:15 here 35 minutes with a bunch of Errors 37:18 of 37:20 course but we see that we we we actually 37:24 did just one prompt all of this this 37:28 like all of 37:30 this it was all using just one 37:35 prompt 37:38 so yeah if again if we had spent more a 37:42 little bit more time in my tool in in 37:45 scop flow trying to add more pages add 37:49 more features add like um all the little 37:52 pieces that um docu sign 37:55 has this would probably 37:58 we will probably get more errors but we 38:00 will probably um get this done easily 38:05 okay we'll probably have more features 38:09 done we could probably um set up the um 38:14 describe better how we wanted the 38:17 document to be signed how to send a 38:19 documents how to 38:23 um we see here that we 38:25 can I'm I'm I'm bit press like we have 38:28 to say that we want to take this and be 38:32 able to place in where that we want we 38:34 probably need a library for this and we 38:37 um didn't specify in any 38:41 dependencies but this is this was pretty 38:45 cool okay we have we had a bunch of 38:48 errors but at the end we were able to 38:51 build this 38:53 okay it's not 38:55 done it's not not been close to be done 38:59 but I loved the uh the laning page and 39:03 also I love the uh this 39:06 um this um off page I I love the design 39:11 this will probably be something that um 39:13 I will try to get a little bit more a 39:16 bit closer to what duck sign is I try to 39:21 get a uh the feature of the document 39:24 done connected to super base and then 39:27 and send to cursor to um like to 39:31 refractor the the the components the the 39:33 pages that we need try to split the 39:37 components so we can load faster and 39:41 um create the um the edge functions to 39:45 like send the document to the um to the 39:49 person that need to sign 39:52 this and have a secur for this person to 39:56 click the link that is a unique link and 39:58 then go to the page and then be able to 40:01 authenticate themselves and then um sign 40:04 the document but this is with 35 minutes 40:08 one prompt and only one prompt this is 40:10 just incredible so Greg I'm sorry I 40:14 wasn't able to uh build the entire Duck 40:19 sign clone in 35 minutes but we got 40:22 pretty close if I had a better laptop 40:25 and 40:26 if uh we didn't had that many errors 40:29 we'll probably um get this done in 40:32 probably like 20 minutes or even less 40:35 but yeah so Greg eberg gave us this 40:40 um this Challenge and 40:44 we almost done we almost done okay um if 40:49 you want to keep uh watching me um 40:53 trying new things and building this 40:57 clones or getting more challenges from 40:59 Twitter and 41:00 Linkedin um and learning more about 41:03 loveable and cursor and other ai2 make 41:06 sure to subscribe to the channel and hit 41:08 the Bell button if you have any question 41:10 leave any the 41:12 comments thanks for watching me and see 41:15 you in the next video