Master Lovable: What expert builders need to know

0:00 in this video I'm going to walk you 0:01 through how to use lovable. deev which 0:03 is another text to web app platform 0:05 similar to bolt. new that lets you build 0:08 MVPs and prototypes with relative ease 0:10 while bolt is great lovable has some 0:13 advantages that make it a lot easier to 0:15 use for certain functions such as 0:16 setting up authentication hooking up a 0:19 database as well as even enabling 0:21 payments through stripe which has been a 0:23 bit painful using something like bolt. 0:25 new up to now I'm going to start with a 0:27 big picture overview of some of the key 0:29 Concepts you should understand 0:30 understand when you're trying to 0:31 differentiate between these two 0:32 applications and when you can or should 0:34 use each of them and then I'll jump 0:36 straight into lovable to start building 0:38 and showing these Concepts step by step 0:40 instead of just going through a bunch of 0:41 slides we'll actually do application 0:43 right away so you can see how to connect 0:45 a database how to enable stripe and how 0:47 to build certain types of web 0:49 applications that you can now embed in 0:50 other websites as well I'm going to 0:52 first Go full teacher mode and pull out 0:54 a whiteboard to walk you through some of 0:55 these differences what level is and 0:57 explain what superbase is the core 0:59 database that it uses to actually enable 1:01 all its functionalities all right so 1:03 teacher mode activated I'm going to walk 1:05 through the difference between lovable 1:07 let's just capitalize 1:10 that as well as bolt. new so let's make 1:13 a little separator here and then we'll 1:15 use 1:16 bolt. new all right so on The Lovable 1:19 side it's built on top of this database 1:21 we'll call it a hypercharged or 1:24 supercharged database it's called 1:25 superbase it's even in the name super 1:28 database and then you have bolt. new 1:31 built on top of I'm just going to call 1:32 this web browser infrastructure so I'm 1:36 going to call it infra I don't want to 1:38 get into semantics and details cuz 1:40 that's not the point of this video this 1:42 is super high level just to grasp the 1:43 core key difference here that is 1:45 responsible for all kinds of different 1:47 functionality differences so on The 1:48 Lovable side you can imagine you have 1:51 this database okay and then we have some 1:54 rows and columns and then on top of this 1:57 database you also have functions so we 2:00 call this let's say function one and we 2:02 have function two these functions are 2:05 called Edge functions okay so these Edge 2:10 functions are what allows you in lovable 2:12 to actually create a stripe integration 2:15 to create a hook to a database to allow 2:18 for authentication so these are let's 2:21 say mini helpers that help the database 2:23 do more than just house data it helps it 2:25 actually be very Nimble and execute 2:28 different workflows from One Core 2:30 environment versus building a whole 2:32 backend similar to what bolt does where 2:35 you are hosting that somewhere and it's 2:37 being executed while you're going 2:38 through the app it's a lot more 2:40 lightweight in the capacity that you're 2:41 using an actual database on the bolt 2:43 side you basically have again with the 2:45 actual web browser it's not built on top 2:47 of a core database now they have added 2:50 superbase integration recently so this 2:52 is as a few weeks ago but it's not 2:54 actually fundamentally designed on 2:56 superbase or a database like superbase 2:59 so the way this creates functions is 3:01 wildly different from the way this 3:02 creates functions meaning certain things 3:05 are easier to do in Bolt certain things 3:07 are easier to use in lovable it really 3:08 depends on the use case you're looking 3:10 for now in terms of the llms that play 3:13 or how the app actually fundamentally 3:15 Works they're very similar in the idea 3:18 that they both have an llm or llms 3:21 actually orchestrating everything so 3:24 with bolt you primarily have Claude 3.5 3:27 sonnet as the llm of choice whereas 3:30 based on the documentation of lovable 3:32 you have open AI you have Gina AI you 3:36 also have uh I think also CLA or 3:39 anthropic so I think there's a hodg 3:42 podge of ones being used in lovable 3:43 versus ones in bolt on you so this is 3:46 one core one that's being responsible 3:48 for the bolt on new infrastructure and 3:50 then you have a midly for lovable and I 3:53 realized I should have made this orange 3:54 but you get the point comparing some 3:55 other features at a high level so both 3:58 of these functions have the ility to go 4:00 back in time that is my poor way of 4:02 drawing that let me do that in blue 4:04 actually so both of them allow you to 4:07 roll back if things get hairy which you 4:09 know all of these things are still 4:10 experimental meaning at some point you 4:12 will hit errors it's almost a guarantee 4:14 if you want to build a prototype or an 4:15 MVP of value so that's one thing they 4:17 share one thing that lovable has that 4:20 bolt doesn't have is it allows you to 4:22 select a specific element so I'm just 4:24 going to push down the marker here so in 4:28 the actual box instead of taking a 4:30 screenshot of your actual application 4:32 saying hey fix this thing you can 4:34 actually click on the very bottom of 4:35 your screen which you'll see an actual 4:37 element detector where you can click on 4:39 the screen and say this part of the 4:41 actual web page I want you to change it 4:43 whereas bolt has a function that lets 4:46 you automate basically the writing of a 4:49 prompt it helps with the prompt 4:50 engineering side a lot more which most 4:53 of the time is super helpful and that's 4:55 a feature that's yet to be in lovable 4:57 itself both of them also connect to 4:59 GitHub with level bolts kind of a 5:02 oneclick button whereas with bolt you 5:03 just have to do a little work to get 5:05 there but it's completely doable they 5:07 both actually use pictures as input if 5:10 you wish so I'm just going to put that 5:11 in the middle here so picture and when 5:13 it comes to using them for specific use 5:15 cases I find that for now obviously 5:17 these things change by the hour and the 5:19 day that bolt. new is better for General 5:23 UI plus ux so in layman's term it's 5:27 nicer on design so has a better 5:30 understanding of user interface and user 5:32 experience whereas lovable has an 5:35 advantage that you can render certain 5:37 elements in a page that are tricker to 5:39 do in Bolt And if that will make no 5:41 sense but imagine you had some form of 5:44 weight list you were setting up and on 5:46 that weight list you not only wanted to 5:48 easily connect some form of database 5:51 right to apply early or to apply right 5:54 so this would be super base so I'm just 5:56 going to put 5:57 Supa but maybe you want to throw in a 5:59 loom video or a vsl so you want to be 6:01 able to embed something on the actual 6:04 web page so imagine you want to have a 6:05 video here right so if you try this and 6:08 I'll show you an actual example of this 6:10 as one of our first use cases this will 6:12 work pretty well on the first try on 6:14 bolt it won't and it does get back to 6:17 how both are designed it's not to say 6:19 this is impossible but it's infinitely 6:21 easier on lovable at least the way it's 6:23 designed all right so that's enough 6:24 yapping I'm going to ditch teacher mode 6:26 now and we'll jump straight into lovable 6:28 where we'll experiment with a few things 6:29 you can gauge these differences in 6:31 action versus me just speaking about 6:32 them all right so before we even write 6:34 our first prompt to create a very basic 6:36 application in lovable I'm going to walk 6:39 through how to actually set up superbase 6:41 which I've been speaking about non-stop 6:43 the whole time so this is the actual 6:45 website again there's nothing crazy 6:47 about the actual app itself it is a 6:49 database one advantage it has over a lot 6:51 of databases is it allows you to store 6:54 vectors and if you don't know why that 6:56 matters if you say want to store certain 6:59 VOR values maybe an input from a user in 7:01 a way that uh an llm can scan and use as 7:05 input vectorizing is essential so one 7:07 key Advantage with superbase is you 7:09 could create an edge function if you 7:11 remember that's a helper function that 7:13 would allow you to take certain columns 7:16 and create vectors out of them which are 7:18 just texts and numbers that allow llms 7:20 to process it and understand them so if 7:22 there is some form of follow through in 7:24 the application you want to create that 7:26 gives you a certain Advantage so if you 7:27 go here to product and you see here 7:29 Vector toolkit you'll see that it 7:31 basically lets you have a mini Vector 7:33 database on top of your normal database 7:35 which is why it's really useful and it 7:37 keeps you very Nimble in terms of the 7:38 types of applications you can actually 7:40 create so when you go in and you log in 7:43 you'll see some blank screen and then 7:45 you'll be able to create your first 7:46 project so for the purposes of this 7:48 tutorial let's create a project and one 7:52 prerequisite is that you'd set up some 7:53 form of organization on the free plan I 7:56 believe you get two different projects 7:57 for free and beyond that you have to pay 7:59 a certain amount per month so let's 8:01 create a new project and we'll call it 8:03 lovable demo and for this I'll use small 8:09 or let's use micro so there is a payment 8:11 here per month uh to be used so be 8:13 mindful of that and then you set up a 8:16 database password so I'm just going to 8:17 set up a pretty simple one here and then 8:20 you choose the region where your 8:22 database should be hosted so I'm going 8:23 to choose Canada since I am in Canada 8:26 I'll create new project and this will 8:28 take around 4 to 5 minutes so I'll just 8:30 wait for this to render and load and 8:33 you'll see here it will say setting up 8:35 project and once it's done you should 8:37 see two buttons on the right hand side 8:39 both that are green that tell you greens 8:41 means go you're good to actually use 8:42 this database all right and it took 8:44 around 6 to 7 minutes I actually finish 8:46 but now you can see project status and 8:48 security issues are all resolved so I'll 8:50 give you a very quick tour of superbase 8:52 on what the features that you should 8:54 care about are so if you go to table 8:56 editor here this is where lovables is 8:59 going to actually create and write the 9:01 databases depending on the application 9:02 you're looking to make so you're going 9:04 to see this populated once lovable 9:06 actually knows what you're trying to 9:07 build when you go to Edge functions here 9:10 this is where a lot of the functions 9:11 that you'll need let's say if you want 9:13 to integrate stripe or you want to do 9:15 different functions that are outside the 9:17 scope of normal backend this is where it 9:19 would be hosted and if you go to Project 9:22 settings and look at Edge functions here 9:26 this is where certain secrets are going 9:28 to be stored so when you set up your 9:29 database for the first time which you 9:31 don't have to do lovable will just know 9:33 hey we need to create some form of 9:34 database because you want to store XYZ 9:37 information about the user or their 9:39 preferences or whatever it's going to 9:41 create some API keys for superbase here 9:44 because it's going to be integrated with 9:45 your infrastructure and then you can 9:47 click on add new secret to enable things 9:49 like openai or other llms to have their 9:51 secrets actually embedded in the app 9:53 itself beyond that if you are logging in 9:55 and setting up an account for the first 9:57 time you're going to want to go to 9:59 authentication and providers and pick 10:02 which providers you want to set up out 10:03 of the box for me personally because I 10:05 like to spin things up pretty quickly I 10:07 just choose email and I remove this 10:10 confirm email and secure email change to 10:12 make things a lot easier otherwise when 10:14 you create authentication in the app 10:15 it'll have to confirm the email and what 10:17 happens is the confirmation goes nowhere 10:20 because you have to set up what's called 10:21 a redirect URL where it comes to here 10:24 this URL configuration so if you're just 10:27 experimenting with this to see what's 10:28 possible I would I just remove those 10:30 features for now just to remove any 10:31 headaches or friction from actually 10:33 getting going so with that we can now 10:35 use this database and we'll create our 10:36 first mini demo just hooking up to 10:38 superbase to see how that works so we'll 10:40 go into lovable here and in here I will 10:44 first click public to private if you 10:47 start with a free account it will be 10:50 public by default meaning if you go to 10:52 featured or latest you will be able to 10:55 access other people's projects there and 10:57 look at their chat history if you did 10:58 want to make sure that all your projects 11:00 are private then you would want to 11:01 upgrade to their $20 a month plan that 11:03 allows you to do that and one thing to 11:05 do is if you want to stick to the free 11:07 just to try this out be very careful not 11:09 to put API keys or anything publicly 11:12 available that could hurt you or your 11:14 wallet just because you can 11:16 theoretically go into any of these 11:18 featured apps right and go through their 11:20 history and scroll up and see pretty 11:22 much everything that happen in that 11:23 conversation which is awesome if you 11:25 want to be able to see some of the 11:26 building blocks and examples of prompts 11:28 that are more effective than others to 11:30 build actual applications but not so 11:32 much if you're putting publicly 11:34 available information such as API keys 11:37 or anything that can hurt your wallet 11:39 that would be ideal so be very wary of 11:40 how you enter certain API keys if you 11:42 use public now if we go back up here we 11:45 will say something very basic so let's 11:48 create an app that allows me to enter my 11:54 top five business 11:57 schools create a design that is 12:03 futuristic 12:06 modern and has moving gradients in the 12:11 background that are minimalistic now I'm 12:15 keeping this very basic just to be able 12:17 to show the functionality so I'm just 12:18 going to click here on private we're 12:21 going to send this and we'll wait for it 12:23 to do his first actions 12:52 and this took around 2 minutes but now 12:54 you have a very basic user interface 12:56 that lets you enter goals so I want to 12:58 make more profit you can see the little 13:00 gradient behind the scenes um let's do 13:03 another one I want to hire many Allstars 13:07 let's add that and I don't think save 13:09 goals actually does anything because if 13:11 we refresh the page because there's no 13:13 database to store this information 13:15 there's no what's called persistence so 13:17 if I refresh this everything will go 13:19 away because it's not tied to a user 13:20 account so this is a good opportunity to 13:22 show you how their native super base 13:24 actually works so if you click on super 13:26 base here and you go down for the first 13:29 time you'll just need to authenticate 13:30 into super base and then after you'll be 13:32 able to connect to whichever database 13:34 you wish we're going to use our actual 13:37 project we put together here which is 13:39 lovable demo I'll click on connect and 13:42 then it pretty much enters an actual 13:44 request that says please connect my 13:45 super based project lovable demo so one 13:48 thing you'll notice about lovable is 13:49 even when you get errors it writes a 13:52 prompt that you can see to actually send 13:54 it to the app to try to resolve the 13:55 issue so you'll see here it says I'm now 13:57 connecting to your superbase pro project 13:59 yada yada yada it says it's properly 14:02 authenticated it stores real data and 14:05 there we go it tells you what kind of 14:07 edge functions you can use and what you 14:09 can do with that so email notifications 14:11 payments Etc so now if we go into our 14:15 lovable backend we still don't have a 14:17 database because we haven't told it what 14:18 to store so in this case I'm going to 14:20 say can we create a 14:26 authentication with email and 14:30 password that will allow 14:33 us 14:35 to log in 14:38 and save our goals by account so again 14:42 we're keeping things very basic here so 14:44 as this is actually generating the first 14:46 time I won't actually cut I'm just going 14:47 to show you even if you're not uh SQL 14:50 Savvy and SQL stands for standard 14:52 querying language which is a database 14:55 language to actually create tables it is 14:57 doing all the work on the front end here 15:00 to create a table called business goals 15:02 it came up with which Fields it should 15:04 store so in this case it came up with ID 15:06 which is some form of random ID that's 15:08 assigned as a primary key that's 15:10 basically the ability to create a very 15:13 unique novel ID for a user that is 15:16 different from user ID which once in a 15:18 while can be um null meaning it's 15:21 non-existent if the user hasn't actually 15:22 authenticated yet and goal text is what 15:25 you'd actually save here once you add a 15:27 goal and then created at would basically 15:29 create a Tim stamp which is basically 15:31 what time and what day was this goal 15:33 created so one thing you'll notice is 15:34 this thing called create policy and 15:36 there's a lot loaded here that you 15:38 shouldn't care about unless it becomes a 15:40 problem later on the only time it 15:42 becomes a problem is let's say you have 15:43 different user accounts and let's say 15:45 it's a matchmaking app of whatever use 15:47 case you need to be able to see and 15:50 remove security at an individual account 15:52 level to be able to see all the values 15:54 from all the accounts in order to do the 15:55 matchmaking um in layman's terms if you 15:58 know user a has four goals and user B 16:02 has three goals because of the security 16:04 the way it's set up by default user a 16:06 can't see user be's goals but let's say 16:08 we wanted to matchmake these business 16:10 owners by their very similar goals we'd 16:12 have to remove this layer of security to 16:14 enable that again you can remove that 16:16 for now until it becomes a problem so 16:17 one thing that lovable does is it tends 16:19 to ask for permission a lot so in this 16:21 case it tells you what the plan is it 16:23 says after you approve the SQL commands 16:26 again SQL is the language that you 16:27 create right to databases it will do the 16:30 following so it'll create a UI a user 16:32 interface with email and password 16:34 authentication it will set up superbase 16:37 to create the actual table and then it 16:39 will do some error handling meaning 16:41 it'll set up some notifications to know 16:43 um if the user hasn't signed up with a 16:45 proper email or they haven't 16:46 authenticated themselves it will have 16:48 different roots to handle those 16:49 scenarios so if we click on apply 16:52 changes most of the time for something 16:54 simple like this it'll say the migration 16:55 has been applied successfully sometimes 16:58 you'll run into scenario where it says 16:59 it there's an error and it'll get this 17:01 little notification on screen that'll 17:02 say fix so most of the time I'll click 17:05 fix two to three times and then I'll 17:07 intervene myself if things are starting 17:09 to get a bit hairy all right so now it 17:10 has the UI with the email and password 17:13 and you can notice the font Choice here 17:15 is absolutely horrific so I'm going to 17:16 use this feature I alluded to before 17:18 called select I'm going to go into this 17:20 box here until it hovers so you can see 17:23 this is a section this is just the text 17:25 and as you go down it'll tell you 17:27 whether or not you're selecting 17:28 something that is editable so I'm going 17:30 to click on this so now it knows I'm 17:32 referring to this part of the screen I'm 17:34 going to say the font is awful please 17:38 make it something that goes with the 17:41 background and ideally throw some design 17:45 on the page with Emojis so let's see if 17:49 that applies properly so now you'll see 17:51 it's edited the core font here and one 17:53 thing I couldn't select was this lime 17:55 green font so let me screenshot this and 17:58 use this as input and say change lime 18:01 green to bold white font I can't see 18:06 anything all right so we got our first 18:07 error here it did change momentarily the 18:11 font but then it said build unsuccessful 18:13 so we have some error here I'm going to 18:15 act like I don't know what it is I'm 18:16 going to click on try to fix it and 18:18 again sometimes this can be a loop that 18:21 lasts two three times sometimes you have 18:23 to step in one thing you can do is click 18:25 on show error and see what the error is 18:28 this might mean absolutely nothing to 18:30 you but worst case you put this into 18:31 something like perplexity and GPT and 18:33 maybe you can nudge along what it should 18:35 do to resolve it but typically for 18:37 something like this it's pretty 18:38 straightforward that it should be 18:39 solvable pretty quickly so I ended up 18:41 getting a few errors here you'll see 18:42 here build unsuccessful build 18:44 unsuccessful so it for some reason 18:46 couldn't realize that I'm talking about 18:48 this specific email address so it 18:50 remained green so I'll show you a little 18:51 hack and I'm happy this ER came up so 18:53 what I'll do is I'll click on open 18:56 preview to new tab and even if you're 18:58 not Tech Savvy you'll see exactly what 19:00 I'm doing here so if you go on and this 19:02 is on Google Chrome you have the 19:04 equivalent of this in Firefox Safari Etc 19:06 if you go to more tools and you go to 19:09 developer tools you can click on this 19:12 icon and by the way this works on any 19:14 website so you can actually look at the 19:15 element of any website if you want to 19:17 emulate it as well so you can see here 19:18 as I'm going down I can see the name of 19:21 certain elements so when I click on this 19:23 it shows me some name so all I did is I 19:25 copied this name and then clicked on 19:27 this Arrow went back found this copy 19:31 pasted it if you see here all I did is I 19:33 need to change this color to bold white 19:36 I just copy pasted what the element name 19:39 is from here as well as the second one 19:42 and then I just said make it white and 19:43 as soon as I did that you'll see here it 19:45 worked immediately so that part was 19:48 alleviated there so now if we close this 19:51 and we set up an account for the first 19:52 time and we'll do this and let's make 19:55 some password and sign up 19:59 so in this case I just want to double 20:00 check that we disabled authentication so 20:03 if we go on authentication and then 20:07 providers I think email went back to 20:09 confirm email so let's take that off 20:11 here and let's we can log in it'll even 20:16 if we didn't confirm the email it will 20:17 be able to log in given how it's 20:18 structured so I'll just click on sign in 20:22 let me try 20:26 this okay so let's let's 20:30 see um let's do 20:35 this let's paste the error what's going 20:39 on here so I ended up getting a few more 20:42 issues here that popped up and I tried 20:43 to click fix and what happened was 20:46 because originally our super base for 20:48 some reason it reset so if you go to the 20:51 the section I said authentication and 20:53 then providers you'll see if you go back 20:55 to email these are all clicked on so it 20:58 did need you to confirm your email and 20:59 because the email was not confirmed it 21:01 was having issues so if I disable these 21:04 now and I refresh just to make sure this 21:06 took 21:06 effect let me just do this one thing 21:09 that I forgot to do is after actually 21:11 toggling these off you have to go and 21:13 actually click save so that didn't help 21:15 my case and just so I can start from 21:17 scratch I went to here users and then 21:21 the two users I made which is my prompt 21:23 advisor email as well as the temp email 21:25 I put together I deleted them from the 21:26 database so they atically if I go in 21:29 here now and I do Mark a prompt advisors 21:32 and I create a password and I sign up 21:35 right so I should be able to log in 21:36 immediately because there's no 21:37 confirming email and now if I add a goal 21:40 and I say I want to make millions as a 21:43 goal and I add that theoretically that 21:46 should update our database in real time 21:48 if it's synced up in the proper way so 21:50 how you can check that is if you go back 21:51 to table editor you'll see now lovables 21:54 created through superbase a table called 21:57 business goals and if if I refresh this 22:00 you'll see that there's nothing here so 22:02 what I'll do is I'm going to ask it um 22:06 so I was able to log 22:09 in but it doesn't seem like my goals are 22:14 being stored at the account level so 22:17 you'll see here in the message you came 22:18 up with it said looking at the RLS 22:20 policies so just think of that as 22:21 security policies I noticed that while 22:23 we have yada yada yada basically there's 22:25 no permission to write and read from 22:26 there so it basically wrote this SQL 22:29 code this database code and I'm going to 22:31 click on apply most of the time they'll 22:33 say successfully so there we go so 22:35 theoretically if I now have this View 22:38 rerender and I add a goal it should add 22:40 it to the actual database let's sign 22:42 back in and then I'm going to enter 22:47 password and we have nothing logged in 22:49 now right so let's do I want to make 22:54 millions whoops let's do add goal and 22:57 click on Save save goals okay goal saved 23:01 so we have an error here so I want to 23:03 see if the errors is associate to saving 23:05 at all so it did save it so the error we 23:08 got was for something unrelated so I'm 23:09 going to click on show logs here um I 23:13 feel like it's just not communicating in 23:15 the correct way but it did the right 23:16 thing so I'm just going to click on try 23:18 to fix it I won't hold you hostage until 23:21 this resolves and I'll show you what it 23:22 looks like after so it looks like we're 23:24 in the clear now so if I say another one 23:26 I want to hire unicorn 23:29 I do add goal I click on Save goals and 23:32 I go back here I don't even have to 23:33 refresh you can see it's actually sync 23:35 towards here so this is working now and 23:37 if we log out our test can be let's sign 23:40 up with a temporary email so I'm going 23:42 to go to Temp hyphen mail.org I'm going 23:46 to create a new account okay I'm going 23:49 to create another password I'm going to 23:51 sign up and then let's just add a goal 23:55 test goal and I click add this should 23:58 save under a different user ID a 24:00 different created ad and it should store 24:02 separately so if I refresh here we 24:05 should have oh one thing I forgot is to 24:07 actually save the goal so that should be 24:09 something I changed in the actual 24:11 interface so for now let's click on this 24:14 you can see here we got test goal for a 24:16 different user ID a different unique ID 24:18 for the actual goal itself so if we go 24:21 back here let's now step this up just a 24:23 tad and say okay now I want to be able 24:29 to 24:30 automatically save the goal 24:33 anytime I enter it 24:37 without 24:39 having to basically click save goals to 24:43 click save goals in order to write to 24:48 superbase this will make the oops 24:52 spelling mistakes here will make the 24:54 user experience more seamless so it says 24:58 there's an autosave functionality now 25:00 which is looks weird it looks like it's 25:02 infinitely saving things to the database 25:04 so I'm just curious what chaos is 25:06 happening okay um so I'm just going to 25:09 take a screenshot here and then I'm 25:12 going to paste it here I'm going to say 25:14 we don't need this to autosave every 25:19 moment like it is now just 25:23 when we actually enter the core goal 25:29 otherwise there's 25:32 useless functionality Autos saving 25:35 nothing all right so now it finished it 25:37 didn't take more than just one little 25:39 reminder and now it said something very 25:41 important here so it said the business 25:43 goals whatever file is getting very long 25:45 so one thing lovable is interesting at 25:47 doing is it looks at all the AI code 25:49 that's been generated and once in a 25:51 while it will tell you hey I think you 25:52 should summarize this down cuz it's 25:53 getting a bit too meaty there might be 25:55 too much fat in your code so you can 25:57 actually say 25:59 sure refactor the code without 26:03 disturbing any functionality and then 26:05 one thing to check here is when we did 26:07 add goal now if we do test goal three 26:09 and we click on ADD goal right this will 26:12 automatically save so if I refresh here 26:14 this should say test goal 23 like you 26:16 saw here so we're good on that front so 26:19 I'm just going to send this here to do 26:20 that code refactoring which again is 26:22 just summarizing and leaning down your 26:24 code to be as Nimble as possible all 26:25 right so now that it's done refactoring 26:27 the code we can say let's now rework the 26:32 underlying design to look more like 26:37 Spotify design but instead 26:41 of green and 26:44 black It's a combination of white and 26:49 the current Violet we have so I'm just 26:52 more so showing you how you can alter 26:55 design by giving it a reference point so 26:56 Spotify is very well known in terms of 26:58 the actual design principle so if you 27:00 send that we should be able to change it 27:02 so with a few different iterations at 27:04 first it gave me a poor attempt here 27:05 where it wasn't the nicest it was on a 27:07 blank page I basically said try again 27:10 make it a bit more sleek and nicer then 27:12 insisted on the white background versus 27:14 the purple so then I said make the 27:16 background purple and the core Parts 27:17 white so vice versa of what you have and 27:19 then we got something like this so we'll 27:21 just settle with this for now I can see 27:22 the font's a bit more Spotify esque um 27:25 let's focus on integrating stripe which 27:27 was a bit of a Frankenstein show using 27:29 it in Bolt if you watch that last video 27:31 I did a few weeks ago if not then we're 27:34 going to go into stripe here and what 27:37 you can do in lovable is say I want to 27:40 integrate stripe so that anyone who 27:45 wants to use this application has to pay 27:51 $1 let's do1 199 a 27:56 month so now what it should do is it's 27:58 smart enough and I think they've trained 28:00 their back end to be able to tell you 28:02 exactly what elements you need to set 28:04 that up and what it's going to do if you 28:06 remember at the very beginning of the 28:07 video we now are going to take advantage 28:08 of an edge function so a helper function 28:11 to go and create some form of 28:12 integration into stripe so here it tells 28:14 you step by step you need a stripe 28:16 account then you need a stripe account 28:18 with a recurring price of $1.99 then you 28:21 need your stripe API key so what I'm 28:23 going to do is I'm just going to go into 28:24 a test account of stripe so I've logged 28:27 into stripe and I've gone on test mode 28:29 which is good because we don't want to 28:31 spend real money on this and you can 28:33 simulate actual payments so if we go to 28:35 payment links I'm going to create a new 28:38 one I'm going to say um business goals 28:43 and then I'm going to say it's recurring 28:45 and then it's $199 per month okay so 28:49 we'll do add 28:50 product and then we need a few things it 28:53 told you you need a price ID so I'll 28:56 show you where to get that when you 28:57 create your product you can go and then 29:00 click on the product itself and then you 29:03 can go click on this number and then if 29:06 you go down you'll see here a new price 29:09 was created if I click on this you'll 29:12 see here the price ID if I click on this 29:15 one you'll see this ID at the top right 29:18 this ID is what we need so I'm going to 29:20 take this and I'm going to go into 29:22 lovable and I'm going to say here is the 29:26 price ID 29:29 okay and then what else do we need we 29:31 have the stripe account we have this the 29:33 API key I think is going to offer to 29:36 embedd it itself so I'm going to say ask 29:39 me for my stripe API key so I can give 29:45 it to you so this is an actual feature 29:46 within lovable that it actually prompts 29:48 you for API Keys instead of you just 29:50 entering it in the chat uh I realized I 29:53 spelled this wrong give it to you okay 29:56 so let me just send this so so it should 29:58 Now log this price ID as our products 30:01 price ID it'll now come back with an 30:04 actual request for the stripe API key 30:06 while I won't show the actual ID where 30:08 you have to go to retrieve it is you 30:09 want to click on developers here and on 30:11 here you want to click on API keys and 30:13 then you'll see a section that says 30:15 secret key you want to copy paste that 30:17 and then input that into lovable so in 30:19 this case I'm going to enter my stripe 30:21 key into this handy dandy button it gave 30:24 us I'll click on submit all right so now 30:26 we get a subscribe to gold tracker from 30:28 the get-go which looks very much like a 30:30 normal software as a service application 30:32 if you click on subscribe now we get an 30:34 error so it's something with the actual 30:37 Edge function so I'm going to let it 30:38 figure that out on its own so I'm going 30:40 to click on try to fix so now we'll get 30:42 this page again and if you click on 30:44 subscribe now within this actual view it 30:46 will just load forever so what you want 30:48 to do is you want to click on this which 30:51 is open preview to new tab you want to 30:53 go here and then you'll get that view 30:55 rendered again and then once you 30:58 actually log in so let's do this let's 31:01 log in right so now I have to pay going 31:05 to click on subscribe now it says fail 31:08 to start checkout process so I'm just 31:11 going to tell it so I opened it and let 31:15 me screenshot it to make it 31:17 easier and then paste this so This 31:20 should lead to the stripe page that then 31:23 I can pay and then it sends some more 31:24 request back to the database saying hey 31:26 Marcus paid so if if I say so I opened 31:28 it and got this error when I clicked 31:32 subscribe now um I think it has 31:37 something to do with the edge function 31:40 after two more tries I'm still getting 31:42 the same error and one thing that I'm 31:43 realizing is because I can't actually 31:45 render this within the view itself this 31:47 little mini view it doesn't really see 31:49 the error so again if we go back to our 31:51 old trick here if we open this in a new 31:53 tab and if we go to the three dots here 31:56 and go to more tools TOS developer tools 32:00 and then we basically try this again 32:02 let's click on subscribe now you get 32:04 this failure you now at least have some 32:06 traceability of the failure so what I'll 32:08 do is I will just take this and I will 32:13 screenshot it and I'll say here are the 32:17 errors 32:19 right it has the same 32:23 error here is the main error and one 32:28 thing here it's invited me to do now 32:29 after running this again and showing it 32:31 that error is it says create checkout 32:33 logs so if I go to create checkout logs 32:36 this is our Edge function for create 32:37 checkout and you can see an error every 32:40 single time so this is the full error 32:43 message and if I copy this it says error 32:47 combining currencies 32:49 interesting um so if I go to this this 32:54 is the core error should only be 32:58 $199 USD per month shouldn't be an issue 33:04 with the currencies so getting the error 33:07 again this is what I'm going to do and 33:08 then this is the whole point of this 33:09 tutorial I'll show you what do you do 33:11 when things go wrong so I'm going to go 33:12 to history here and then I'm going to go 33:16 to right after we did the purple to 33:19 White authentication thing sorry purple 33:21 to White redesign of the app so let's go 33:23 to that I'm going to click restore and 33:26 now we've reverted to 33:28 before chaos ensued and I'm just going 33:31 to go in here I'm just going to delete 33:32 some of these older subscriptions so let 33:35 me deactivate this because I think maybe 33:38 my test account is malfunctioning with 33:41 all of these different uh actual payment 33:44 links so let me just deactivate all of 33:46 them and then we'll create one from 33:48 scratch and just for Simplicity we'll 33:50 make it Canadian dollars because I don't 33:52 know if it really wants me to be 33:53 Canadian so we'll create a new one we'll 33:55 just call it uh add a new product 33:58 goals I'll just make it recurring $1. 34:01 199 Canadian a month and I'll click add 34:05 product and then I will just create this 34:08 link and it'll do the same thing um I 34:11 will just go to goals I'll go to the 34:15 actual price I'll go to Price ID I'll 34:19 click on this I'll take that link and 34:22 I'll go back and I'll close this and 34:24 I'll say Let's uh make it so that any 34:27 anyone who wants to use the 34:31 app has to pay in order to use it ask me 34:37 for my stripe API 34:42 key and here is the price ID okay we'll 34:47 paste that so we'll redo that part and 34:50 I'll go into my developers thing again 34:52 and get an API key here we'll go to 34:55 stripe secret key what add it and we'll 34:58 see if this makes things a little bit 35:00 easier so now when we take our second 35:02 attempt here I'm going to open this in a 35:03 new tab we're going to get the Gold 35:06 Tracker now when you log in you're going 35:07 to see subscribe now which should take 35:10 you to an actual account here and you 35:12 should be able to subscribe so I'm going 35:14 to use the fake stripe credit card 35:16 credit card test so then I will put this 35:21 424242 and let's just go into here to 35:24 make sure we have the right CSV so let's 35:27 do copy and then let's paste here let's 35:31 make up some random dates some random 35:34 numbers and then I'll just put test and 35:38 then some postal code and we should be 35:40 able to click subscribe here and then 35:42 it'll quote unquote go through because 35:44 it's a test function and once that test 35:47 function actually functions we should be 35:49 able to log in and actually use the app 35:51 which you can see here so that's this 35:52 part and I did want to show you one more 35:53 use case before we sum it up so while 35:55 everyone goes over how to build end to 35:57 end applications using things like 35:59 lovable or bolt or wind surf what people 36:01 don't talk about is that you can build 36:03 components with these services that lets 36:06 you embed certain elements in your 36:07 existing web pages so let's say you're a 36:10 small business owner and you want to be 36:11 able to create some form of calculator 36:13 to help with lead magnets or lead 36:15 engagement and let people calculate the 36:17 value of whatever it is that's relevant 36:19 for your business you can actually 36:21 create this mini embedded app in lovable 36:24 and then be able to use it as an element 36:27 in your existing application so we're 36:29 going to create a very basic calculator 36:31 and I'll show you how we can quickly 36:32 embed that and it'll give you some 36:33 inspiration for what else you can use 36:36 these text to web app applications for 36:37 so let's say you have a cleaning 36:39 business and you want to be able to 36:40 generate a rough estimate for how much 36:43 cleaning per week would cost for 36:44 someone's house given the number of 36:46 bedrooms the number of amenities Etc and 36:49 they can do it self- serve so let's go 36:52 here and let's say I want to create a 36:56 very um 36:58 engaging 37:00 calculator for cleaning business for a 37:03 cleaning business um it should 37:08 take the number of 37:11 bedrooms number of 37:14 bathrooms uh square 37:18 footage of the home um and be 37:23 fairly simple for 37:26 anyone to 37:30 use and then when someone clicks submit 37:35 it will then send all that information 37:39 via lead form and realistically The Next 37:43 Step here would be to create it via 37:44 superbase like we just did and log that 37:46 in but I won't prioritize that since we 37:48 know how to do that I'll just worry 37:49 about the form itself um I'll just say 37:52 don't worry about making 37:55 the calculator 37:58 submission um usable or 38:03 functional create the shell where you 38:07 can enter things 38:10 and select from 38:13 dropdowns Etc so we'll just click on 38:17 private here and then we'll click Send 38:20 and then this is going to now send that 38:22 request and ideally create that shell to 38:24 show us exactly what that would look 38:25 like all right and here's the first 38:27 version of the draft of the calculator 38:30 on the left hand side you have like we 38:32 said the number of bedrooms bathrooms 38:34 square footage type of service you're 38:36 looking for to calculate the quote if 38:39 you want a custom quote or something 38:41 more detailed then you can put your name 38:43 email and number again I don't care to 38:45 make this functional because I want to 38:46 show you how to create what's called an 38:48 iframe so you can actually embed it on 38:50 an existing website so what I'm going to 38:52 do is I'm just going to publish this and 38:54 I'm going to click on private here I'll 38:57 click on on deploy this will generate 38:59 some link I don't care to make it a 39:01 custom domain yet what you can do is you 39:03 can go to a custom domain go to netfly 39:06 buy an actual domain and attach it but 39:07 we don't necessarily need to do that if 39:09 we click on this now this is the new URL 39:12 that we have so this is important 39:15 because this will be needed in our 39:17 iframe that we need to embed in a 39:19 website so now we can actually go and 39:21 ask can you create an iframe of this 39:25 calculator 39:27 I can embed on my website and what this 39:32 should do is create some very specific 39:35 code that you can copy paste into a 39:37 portion of the web page to actually do 39:40 the integration and here we get the code 39:42 it says if frame at the beginning and 39:44 end so I know it's the right type of 39:45 information in SRC it has some 39:48 placeholder for website that placeholder 39:51 is this so we're going to need to put 39:53 that so I'm going to just make my life 39:55 easier and say this this 39:59 is my deployed 40:02 URL can you tailor the ey frame just to 40:07 be extra lazy here so it should now make 40:10 another version of this that just has 40:11 that embedded URL just created input it 40:14 in that iframe object all right so what 40:15 we're going to do is we're going to copy 40:17 paste this and I see it added one 40:19 component here that immediately is 40:20 jumping out to me so in case you see it 40:22 on your end make sure you remove it you 40:24 want to remove this slash embed because 40:25 it's pretty much meaningless we just 40:27 just need the actual full URL of the app 40:29 so if you put this in here which is 40:31 html-only 40:35 docomo this embed component and then 40:39 once you do that it should be able to 40:40 actually render and you can see here on 40:42 the left hand side you can see the 40:44 actual screen or if frame that we 40:46 created with that core calculator where 40:48 you can actually click and interact with 40:49 it so let's say we actually made a web 40:52 page from scratch so let's go to chat 40:54 gbt I'm going to use my Pro here I'm 40:57 going to say create a beautiful um HTML 41:01 CSS for uh a web page for a cleaning 41:07 business you don't need to do this on 41:09 your end this is purely just to show you 41:11 how this would work assuming you had the 41:12 website so it finished actually creating 41:14 the website and if we copy paste this 41:16 into the HTML online you can see on the 41:17 left hand side here you have some 41:19 formulaic cleaning page okay so I can go 41:23 back with some quick feedback to chat 41:26 GPT I'll be like hey um I don't like 41:31 this that's all good some picture here 41:34 is not rendering we don't need that um 41:38 I'll say please make the landing more 41:43 minimalistic and remove the non-existing 41:47 image here and I'll say please provide 41:51 me with 41:54 downloadable files to see this local all 41:57 right so now I created an HTML file that 41:59 we can copy paste as well as a CSS file 42:03 if you don't know what a CSS file is 42:05 it's pretty much responsible for all the 42:07 styling of a particular web page so 42:09 think of icons colors fonts Etc so we 42:12 can copy paste both of these into two 42:15 separate files so that's what I did um I 42:18 created a new folder called uh web page 42:21 and then I called one index.html and the 42:24 other one like it said uh styles. SS 42:28 it's important you name them these 42:29 specific names because they reference 42:31 each other so if you name this Styles 2 42:35 without updating certain components of 42:36 the HTML to refer to Styles 2 then 42:40 they'll not be able to talk to each 42:41 other and then you'll have just a one 42:43 page full of links and hyperlinks that 42:45 looks pretty ugly so what I did next is 42:47 now that I have them both in one folder 42:50 I use this app called brackets it's free 42:52 to use it basically lets you basically 42:54 render things like HTML or css on your 42:57 local desktop so if I click on this bad 43:00 boy you'll see the actual web page as we 43:03 saw it before now it looks a little bit 43:05 more minimalistic and what we want to do 43:07 now is go back and the whole point of 43:10 this is to take this and then I want to 43:13 ask it where do I put this iframe to 43:17 embed it in the middle center of my web 43:23 page and then we'll put this here and 43:25 I'll just remove the embed like like we 43:27 did before and then it should give me 43:29 some updated HTML not CSS necessarily to 43:33 be able to make sure that embed actually 43:35 appears on that web page all right so 43:36 initially chat gbt gave us small 43:38 Snippets of code where it said hey take 43:40 this and put this wherever you want 43:42 let's say we don't want to deal with 43:43 that headache I just asked it to Output 43:45 the entire HTML and CSS file so I can 43:48 just copy paste so it gave us these 43:50 again I just copied each one went into 43:53 brackets and now I just copied and then 43:55 pasted same thing with the HTML 43:57 I saved and then once I click Refresh on 44:00 the web page itself you can see now that 44:02 we have our normal web page and we have 44:05 this embed from lovable you can see here 44:07 we remove that and theoretically you can 44:10 make this into one page you could extend 44:12 it you could play around with how it 44:13 appears on the web page but the bottom 44:15 line is now you have a smart app that 44:17 you can even integrate gbt 40 or gb40 44:20 mini if you want to add some natural 44:22 language components that you can embed 44:24 on your existing website so the whole 44:25 point is just to show you you don't have 44:27 to necessarily build something end to 44:28 end if you don't want to if you already 44:30 have something if you have an existing 44:32 web application you have an existing 44:34 website you can also use lovable to 44:36 create components that make your 44:38 services that much more helpful 44:39 especially if you want 2025 to be a year 44:41 of more leads more interest and whatever 44:43 you offer this is a good way to actually 44:45 get around that again without having to 44:47 hire a web developer necessarily to do 44:49 the thing for you and while I can keep 44:51 building tons of different use cases on 44:52 lovable what'll give you in the meantime 44:54 are two different things to help you 44:56 Kickstart using it or if you have 44:58 already started using it and you've been 44:59 comparing it against bolt this will help 45:01 you at least understand the nuances so a 45:04 lot of the things that I mentioned today 45:05 such as super Bas stripe um how this 45:08 works compared to other Tex web app 45:11 applications I have this full guide for 45:13 you that will be downloadable in the 45:15 gumroad link in description below and as 45:17 usual in Mark fashion I put together a 45:19 GPT to help you as a co-pilot so like I 45:22 said before there are some nuances in 45:24 prompting lovable compared to other text 45:26 to web Builders so the idea here is that 45:29 similar to my bolt. new GPT I put 45:32 together more than a month ago you can 45:33 go on frame my project and you can say 45:35 something like I want to build a web 45:40 application that takes people's goals in 45:45 their 45:46 business and create a full road 45:50 map 45:52 who for them right and then this will 45:56 try to create an application for you 45:58 like the actual what the look and feel 46:01 design it thinks is a smart idea any 46:03 features you might want to consider and 46:05 if you want to ask it hey can you create 46:08 a succinct prompt for me I can enter 46:12 into 46:13 lovable then it'll do that as you'd 46:16 expect so there you go create a web 46:17 application that generates business 46:19 schoal road maps and it goes through the 46:21 different features and then you can 46:23 actually ask it questions like should I 46:26 integrate super base at the 46:30 start or 46:32 later and it should have some logic from 46:35 what I fed it so it's highly recommended 46:37 to integrate superbase at the start of 46:39 your project here's why and goes through 46:41 some of the benefits and even gives you 46:43 some boiler plate code in case lovable 46:46 hallucinates for whatever reason but 46:47 this should be really helpful for you to 46:49 get started or imilate what you already 46:51 done in terms of your prompting capacity 46:53 and if you wish to take apart my GPT and 46:56 customize it for what you like to build 46:58 on level you can just say what are your 47:01 custom instructions let's open a new 47:03 chat here and you click on oops new chat 47:07 and then I'll just do this and say 47:10 output in a code Block in markdown you 47:15 can see all of my underlying 47:16 instructions you can edit them create 47:18 your own custom GPT of your choice 47:20 whatever makes it easier for you try to 47:22 load it as much as possible but maybe 47:24 the examples you want to tailor to your 47:25 specific use cases feel free to do that 47:27 it's all yours it doesn't hurt me 47:29 whatsoever so enjoy if you found this 47:31 video helpful let me know down in the 47:32 comments below I know it takes a bit of 47:34 trial and error with these errors but 47:36 once you get past them you can now build 47:38 things that otherwise you would have had 47:40 to Outsource to a full stack developer 47:42 and you can really prove a concept or a 47:44 prototype out pretty quickly and even 47:46 collect payment using lovable. deev I'll 47:49 see you all next time