How to Scrape LinkedIn Jobs Automatically with AI (make.com)

Hi, imagine this. You're job hunting on LinkedIn, but instead of manually clicking through each listing to figure out if it's a good fit or not, you have all the jobs neatly organized in a Google sheet automatically updated every few hours. Even better, based on your resume, it shows you a job match score, so you can skip the noise and focus only on the opportunities that truly matter. That's exactly the automation we are building in this video. If you'd like us to set this up for you, check out the links in the description. Okay, let's dive

in. Okay, so here's what we'll cover in this video. Okay, the first thing is that we will start by navigating to LinkedIn and collecting job URLs based on our criteria. Okay, the second is we create an RSS feed to automatically scrape jobs from those URLs. The third is we'll build a make.com automation to scrape job details from LinkedIn and then add them to the Google sheet. And finally the fourth is that we will upload a resume in this case my resume and we will write a prompt in chat GPT which is going to extract the

job description. Then it'll match together and generate a matching score. Now this is important basically because it makes it easier to decide which jobs to apply for. Okay great. So the first thing is that let's go to LinkedIn and we will start scraping the jobs from there. Okay. So let's go to LinkedIn. I'm just going to type that. Okay, LinkedIn right here. Okay, perfect. So, we are here. Now, we'll navigate to jobs and within jobs, I am going to go to show all. Okay, so let's put a criteria here. I'm going to say artificial intellig.

Actually, I've already done some recent research. So, let's say this is artificial intelligence. I'm just going to write it out here. Artificial intelligence. Okay. And then I'm looking for artificial intent jobs in Sydney. Let's put some criterias here otherwise there'll be lots of jobs. Date posted. I want something which is in the past week. There are still 946 results. Experience level I'll say director executive mid. That should be okay. There are still okay. One second. Okay. And you can you can actually make it more if in case you want to apply easy apply or you

know add hybrid and everything. So but for the purpose of this demo I just want to show you exactly how to create a URL. Okay, because the first thing is that we need to have this URL. So this particular URL is right here. Okay, and it has all these jobs. So we've already skipped that and you can create it for uh you know you can create multiple URLs for multiple jobs. So like for if I were to apply for this thing, what I'm going to do is I'll apply for artificial intelligence jobs and then I'm

going to do one for Sydney, uh one for Melbourne if I'm looking for jobs in Melbourne, one for Brisbane and things like that or maybe one for New York or you know wherever you want to go. So I'm looking for jobs in Australia. So this is what I'm going to do. Okay, artificial intelligence and Sydney. Uh this thing another thing which I would be interested in is product management. So that is my background as well. So I will say product management or product manager roles. Okay. And same thing out here as well. So you can

change this thing. All right. So anyways, let's go back and I'm going to show you first um how to do this in artificial intelligence. So we are right here. Okay. So I've put in the criteria. I'm just going to copy this this URL. All right. So the next step is that now we have this job URL. And as I said before, you'll have if you're applying for jobs in Australia, within Australia, sorry, in Sydney, in Melbourne, in Brisbane, you will have three different URLs for that. All right. And next is where we go to uh

we'll go to rss.app. All right. And uh I've already logged in here, but it's pretty straightforward to create an account. Let me just show you. This is the one. This is the homepage. And what it does is basically it creates a feed of the URL you provide. Okay? And that's what it does. So, and um so you can create an account. I think in order for you to create a feed, you have to take a paid plan, which is pretty cheap actually. It's not that expensive. I think we paid something like uh let me just

go here real quick. Okay. So, these are the plans out here. I'm on a basic plan right now because I just wanted to demo. And um up to 24 post per feet. In these cases, you'll get 50 uh odd. So I I think for most uh for this particular purpose I think a basic plan would suffice but of course I mean if you have different requirements then you can go for developer as well. Okay. All right. So we are done with this uh let me just go back to I'm going to close this. Yeah. Now

we have this URL right and uh this particular URL which I've copied and then I'm going to create a feed. So creating a feed is pretty straightforward. What you need to do is just come and paste your UL right here. And then it will just that's it. Click it. Finished. And now it is generating the feed. Okay. Let it just get it done. It should take about probably five or 10 seconds more. Okay. And there you go. So now it just says quantum hiring executive managers um you know and um capability in Sydney, Australia and

things like that. Okay. So it is you can see all the jobs are right here. Now what I'm going to do is I'm just going to say save to my feed and it gets saved to my feed right here. You see this? So I've already created these six you know I don't want to waste time. So I've created this uh product manager for Sydney which I was talking about Melbourne product manager uh this is Brisbane and this is for artificial intelligence for Melbourne, Sydney and Brisbane. Okay. So let me just configure this and I'm going

to show you how to do that as well. Uh so what you the first thing you'll do is you will actually go to filters. Once you come here you uh since it's a feed you don't need any uh you know you don't need any description sorry uh you need a description but you don't need any images. So we remove all that. So apply filters to autohide post with no description no dates no duplicates sorry duplicate titles and duplicate descriptions. Okay. So first put four of these out here. Go down. Anything older than 7 days really

doesn't make any sense, but if it does for you, uh, feel free to actually increase this, but I'm going to put in 7 days right here. Okay. And that's it. So now the feed has changed a little bit. And uh, another thing which I'll do is I'm just going to go to feed output customize this. And then I'm going to remove any images because you just want it straightforward. Uh, title tags, description tags, you don't want don't include thumbnails and don't include videos. Okay. And that's it. Save this and your feed is ready now. All

right. So this is step two. The first one was basically to go to LinkedIn and find the URL I mean job URL and the second one is to create a feed. Now this is how you create a feed. You can actually rename this as well. Uh uh just go to edit and out here we can rename this. So that is what I did. I have created these four fields. Okay. I'm just going to delete this now. Delete feed. Okay. So I've done this for a IML. So if you can see out here, this is, let

me just show you. I'll edit this and this is top 75 artificial intelligent jobs in Melbourne. Okay, same thing for uh jobs in Sydney. So if I go to edit, it'll show me top uh jobs 153 artificial intelligent jobs in Sydney. Okay, and so forth for Brisbane for product management roles in uh Brisbane, Sydney and Melbourne. Okay. Okay. So now we have done two things. One is that we have actually got the URLs from LinkedIn and the second thing is now we have created the RSS feed for those as well. Okay, all the URLs. Now

the third thing which we need to do is we need to actually go to make.com scenario. We need to create a automation scenario out there uh so that we can actually give this feed and it can scrape all the jobs and then kind of process those. Okay. So let's get go to make.com and for those of you who don't know what make.com is, it's an automation platform. Okay, it just uh helps you um it helps two different apps or two or more actually different apps talk to each other. So let's say if in case you

want uh you know to use Google Sheets uh something happens in Google Sheets and you want to send an email out uh from Mailchimp, you can do that via using u make.com and of course for much complicated scenarios as well. Let me just show this to you right now. So this is make.com you if you don't have an account it's pretty straightforward to open an account. All right, so I'm going to go ahead and log in. I already have an account uh and let me just just let it load up. Okay, perfect. So now go

to scenarios. Scenarios is where you build your automation. Okay, so in make platform you call it scenarios. So out here I'm going to just uh create a new scenario. So you can you can create one out here but I already have a scenario already done that. So I'm just going to take you through that. I'm going to take you through each enemy module. So let's go to this particular one active. Actually, it's an active scenario. Okay, this one right here. All right. Okay. So, let me just open that. I already run it a lot of

times. So, let me just take you through this. Okay. So, this is the scenario which I've made. Okay. So, what it does is it actually takes in your RSS URLs which you created and then it will uh you know add it to Google sheet and then do a lot of processing and um you know finally we'll have the jobs in Excel. So let me just show you exactly how it look. This is an example. Okay. So it will get a job title, company, location, post date, job link to this particular one. So if I click

this, it takes you directly to that job as well. Okay. And uh then you have a summary. Uh so what exactly this job is about. There's a summary for that. There's a math score which I spoke about and for this we use chart GPT. And of course we will upload the resume as well. And uh what this prompt does is it takes your resume and it takes the job description and it kind of generates a score so you know exactly how well you match the job description. So it's kind of easy right? Uh you don't

want to apply to every job which is there and uh then of course it gives you the rational as well as to why it gives you a score of 80 based on your resume. So it give you some explanation if you want to go through that. All right. And then if you're interested, I mean I just have two different columns called interested for internal tracking as well as whether the resume is generated or not. And that is something which I'm going to speak about in the and later on. Okay. Okay. So this is how it

will look. It will go through all the jobs and go this is just an example. Of course it's not going to be two jobs. It'll be much more than that. But I just wanted to show you exactly how it look like. Okay. All right. So now let's go back to Mick. All right. So as I said we already have six URLs uh which we have from the six feeds right. So 1 2 3 4 5 6. So we have six feeds and uh in make and what we need to do is we will actually extract

all the jobs from each of those feeds. All right. So now let me just take you through the modules one by one. Okay. The first one is you can see a clock here. Okay. And you'll start the the whole automation sequence by adding this at regular intervals. You can say 240 minutes or you can change this. All right. So this runs every 4 hours. You can make it run every 15 minutes, every 10 minutes, whatever your case is. But remember that every time it runs, it of course uses consumes a lot of uh you know

uh tasks or uh I don't know what do you call them actually. Uh I think uh a lot of transactions I think. One second. What do you call them man? Operations. Sorry. Okay. Yeah. So every time it and you can see that you know it's run. So every time it runs it actually consumes 351 1,228. So there was probably some issue out there. So it it consumes a lot of transactions. Now if you have a big plan it doesn't matter but I'm just saying that you know there's no reason to run it every 15 minutes

because you know jobs are not coming immediately. Uh every 4 hours 5 hours perfectly fine. Within a day I think you should run it about two or three times. Okay. So then you can change everything from here. The next thing is what you do is you add a module to it. Okay? So you can just click here. You can say on this link and just say add a module. Okay. So when you add a module you can choose whatever you want. Okay. So I'm just going to delete this module and uh let's get it back

to this side. All right. Now the first module I chose is a JSON. Now the reason I chose this JSON is because we have six URLs, six feed urls, right? And we want to go through that. Now, one option was that you create six scenarios each with an independent RSS feed because an RSS feed can only take one URL. Okay? But what I wanted to do was I wanted to cycle through all of them. And in case your case, if in case you have 10 URLs or 50 URLs, whatever the case is, you can use

that. So the first thing we need is we use a JSON parser. All right? So out here when you go, you can say JSON parser and then you have JSON built-in. When you click on that, just go to parse JSON. Just click this and this. This is the scenario which I'll be using. All right. All right. So when you open this, this is a string and uh just just remember that this is the way you need to do it and it takes an array format and of course this is the URL one URL 2. Now

if you don't know how to do it, I generally prefer uh when you're doing it like this, just copy this. Go to something called JSON editor. Okay, JSON editor. It's pretty straightforward. So when you click here it will tell you okay whether the JSON is formatted or not. Let's say for example if you're missing a quote or something it will start giving an error saying that there is something wrong with this JSON. This is not parsed correctly or this is not created correctly. So then you can actually use that. Okay. So as you can see

this is an array. Array is nothing but a collection of things. In this case you have a collection of URLs. So there's URL 1 2 3 4 5 6. Okay. So that's it. I'm just going to close this. So I've you can copy that and you paste it out here. So what essentially what this is going to do is this is creates a JSON and this is an iterator. Iterator is nothing but it cycles through. Okay. So in this case we have six URLs. So an iterator basically will go through it takes a JSON array.

Okay. So the same thing when you click on that you will get this option for JSON array. All you do is just pick it up and put it out here. That's it. So the iterator goes through each one of the um you know the values basically. So in this case we have six URLs. So six values. So it'll iterate through those and then it will feed that into the RSS. So again RSS is right here. You can go here and then you can see RSS in my case. Otherwise you can just say RSS and it'll

come up. Okay. So when an RSS module comes up you'll say RSS feed items. That's it. This one is what you'll use. Okay. So if you go and zoom in, you'll see retrieve RSS feed items. All right. So once you do that, it needs a URL. In our case, where we going to get the URL from? From the iterator. Okay. And I'm going show you how it works. Actually, let me just uh unlink this and okay, and show it to you. I don't want the whole thing to run. So let's run this particular scenario only

right now. And then I'll show you what exactly happens. Okay. Okay. So when I run this once and you can see you see that. So this is it ran once. All right. And this is the input. The input bundle is like items in the array 1 2 3 4 5 6 all of these six ones. So what is the output of the array? It is nicely formatted this 1 2 3 4 5 6. Now the iterator basically will go through this each one of them. So this is item number one number two number three and

so on so forth. So let's go to the output of iterator. And you can see this is the output of iterator bundle one. This is the value bundle two. This is the value. Okay. And this is what it's going to feed into the RSS. Now in the RSS you have operation one. So it's already got this bundle and it is got all the output. So it has got remember we can only scrape uh 25 jobs at a time. Okay. So you can see probably it has 24 or more. If there are more jobs then uh

it will come later but and you can see there are 25 for this then you can go down to operation two. This is the second URL. Okay. This is the second input URL and then the output for that particular one again is bundles probably 25 bundles again. All right. Same thing uh for URL 3 there are 25 bundles I think. Just one second. Two it's 20 bundles in this case. Okay. Okay, so there are only 20 jobs and this is probably brispin anyways. So you get the idea. Okay, so this is how you will um

you know this is how we do it. So what's going to happen is that when we run this particular scenario or when it runs every 15 hour sorry 4 hours when it runs it'll come here it'll pick up the same URL it'll go here the iterator picks up the first one and it goes here and then it extracts all those jobs from the RSS feed. Okay, as you just saw, the next thing is what we need to do is we will create a Google sheet. Now, you saw this Google sheet. This is the one which

we've created. Okay, but we don't want duplication. All right, so uh because otherwise what's going to do is uh there is no operations which is going on the RSS. What is going to do is going to just take the same jobs and keep on posting it and then you'll have a lot of duplicate records. All right. So for to avoid that what we need to do is we need to actually add a Google sheet module which again I went here and then I said Google sheet and there is something called search rows. All right. So

this is the module which I've used. So search rows. All right. So let me just connect that. All right. So let's go through this module as well. So again you need to have a connection. Um you know so I'm using my own connection sep.ipites and uh then you have a search method. So when you start search method is searching from a drive when you have spreadsheet. So this is my spreadsheet in your case it'll be your spreadsheet or whatever spreadsheet uh you want to use. So I'm using job board sep which is right here. Okay.

And then sheet name is jobs. If you go down to sheet names right here it is jobs. In your case if it's sheet one or something else then make sure to select that. Okay. And uh table headers. Yes there are table headers. Columns are there. And of course job link. Now what exactly this means is that uh I want to put a filter. If this job uh link is equal to the URL then there the filter is on basically. So what it'll do is it'll see if this exists. Okay. So job link is the fifth

field. All right. So this is a unique job link for that particular job. So we will compare whether this uh job already exists in the sheet and if it exists in the sheet then it will filter it out. Okay. So that's what we do. This job link is equal to the job link which is coming from the feed. Now let's see this particular feed if there is a job link and you have a URL right here. Right? So that's what we are using here. Okay. In the this thing if the text operator is equal to

this thing then you put a filter. And what is the filter? Job link does not exist. So only then it will pass. If the job link exists then it does not exist. So basically it should not exist. If it exists then it will the filter will block that particular job. If it doesn't exist it will pass through and let it go. Okay. All right. Now after that so once we run it basically the first job comes like this. So let's say if there is nothing in the sheet right now there are two jobs and I'm

going to delete that or maybe just let it be doesn't matter. So when the job is coming in if doesn't exist in the sheet it'll start adding the values right here. All right. And uh it'll pass through. So the first one is that what we'll do is we will use a chart GBT module. Now look at this module. This is transform text into a structured data. And why are we doing that? Because we want the job to be coming like this. We have a job title. We want a company location. Okay. The way it comes

in RSS feed is like this. So if you see the bundle and uh the job summary and uh one second this is the bundle. Okay so the title comes like this. Okay it is saying as alois solutions hiring RPA up developer in Melbourne Australia. So it has got this title has everything. It has got the job description sorry uh it has got the job uh company name. It has got the job title as well as it has got the city where it's happening. Now if it comes like this it's okay if you want to read

like this but I prefer you know that it's much more cleaner. So that's what I did. I used so I've created three columns job details uh sorry job title company and location and I wanted to be segregated like this and that is the reason I'm using a chat GPD module. So basically what I'm telling this module is that you will get the data in this particular sequence uh in this particular format and what I want you to do is I want you to break it down and into these three categories which is break it down

into job title, company and location. So when you get a title so this title is coming from right here this one. Okay. So what it's going to do is it's going to break it down into three the things and then and then you have a parameter right here. So that is the structured data definition parameter name you have a title job title and type is text then the second one is company that's again a text and the third is location which is again a text okay and that's it. So once you do that what's chart

GPD this module is going to do is it's going to take that URL and break it down and then uh you know you can use it later on into u adding it to the sheet. Okay now this module what is this? This basically says ignore if there is an error go to the next one. Ignore the error. Okay. So, and how do you do that? Um, you don't have to add it basically, but I've just added it. If you want to add this one, you'll go to flow control. Uh, this is flow control with flow

control. Okay. So, you have flow control and this is something which is inbuilt in make and then you can use this particular one right here. Uh, ignore. Okay. That's the one which I've used this particular one. Okay. Okay. So this is done. So now by this time we've taken the sheet we have searched the roles only the filter ones are led to pass through. Once it comes in the chart GPD the first module actually breaks it down into title uh job title company and location. Okay. After that what we do is I need to format

the date because uh there was the date was not coming right you know. So this date it was coming in a very different format and I want a post date and a time as well. So what I did is I've used a tool uh this is and then there's a function within that called format date. There's a internal function to make. So if you do format date let's go to date somewhere around here. General functions I think was format number maybe it's in the date column. Format date is right here. Okay. So this is what

I've used this one. And once you do that the date format gets uh you know in it comes in a proper one so that it can be added directly to the sheet. Okay. So once we've searched it and now this all this is coming now we need to add it to the sheet. Okay. So out here again I've created a Google sheet module and it is saying add a row this one. So again go back here to plus sign go to your Google sheets and within that you have add a row just add that. So

right here okay I'm just going to delete that and then you can connect it. All right. So when you click this again you have to select the connection and once you click the se the connection you have to go and uh go to the sheet which which sheet uh you want to use. So in my case it's Nvidia resources AR resumeum builder for LinkedIn and job board separates with the values of the sheets. That's what it picks up all the headings. So the job title, job company name, uh location and then post date. So I

have a lot of these things and then I have to fill them up. Okay. So now where is this value coming from as we said right? So this is coming from the title and right here. So whatever is in the title it whatever is coming in the uh in in the chart GPT is going to go in the title. So I've used this as you can see right here. Title, company, and location. That's what it's broken down to. So you will just pick it up and put it out here. Okay, that's it. So title, company,

and location. And then the date is already formatted. So the date was not coming properly. So I formatted that as well. And that you just pick up and put it out here. So now we have the job title, we have the company, you have the location, you have the post date and you have the job link. Now the job link is directly coming from the RSS feed. You don't have to do anything with that. So it is coming directly. If you go down, you can see the link right here. So you use this URL. Okay,

that's it. All right. So, yeah, you must be wondering why I didn't use this date. But if you use this date directly, for some reason the formatting goes away and uh you know in Google sheet. So that is the reason I had to use the function uh the formatted function. All right. So I've done this now and uh the Google sheet as you can see it'll get filled till here. Okay. But I've not generated the summary match rationale and all those things. So for that what we'll do is we will go again to tools. The

next one is set a variable. Now as I said before, now this is step four. Okay, so till this time we have extracted all the the the job basically from the RSS feed and we have populated the sheet. Now if you want to see, let me just show it to you. Let me just unlink this for for a minute and I'm going to run this. Okay, so basically what it's going to do is it's going to populate that. Okay. And uh I'll delete it later on. But let me just show this to you. So this

is unlin now. Or is it linked? Oh, this is unlin. Okay. So it'll go only till here. So let me just run this now. And what it's going to do is it's going to read the first URL and go all this and then start adding it to it. So let me just go here. Okay. Let it be out here only. That's all right. All right. So run once and you can see one and it's passed through and it's added. Then again next two two added and let's go here. See it's adding the thing right. Okay

perfect. This is what I wanted to show you. Um all right just how do I stop it? Stop for stop wash. Okay perfect. Just a quick note if you're enjoying these videos don't forget to like and share. It really helps the channel. And if you're into automation content like this, hit subscribe so you never miss out when I post a new video. If you're new to automation, I highly recommend joining our AI accelerator. You learn how to build cuttingedge AI systems for your business or clients, plus connect with like-minded people to boost your motivation and

success. All the links are in the description. Lastly, drop a comment if you would like me to expand on this particular automation. you know, like automatically generating tailored made résumés based on the specific job descriptions or if you have any other cool ideas to enhance this workflow. I'd love to hear your thoughts. All right, let's get back to the video. So, this is what happens. Basically, it starts adding it up. Okay, but it doesn't have the summary. It doesn't have a math score. It doesn't have a rational and all that stuff. And I wanted to

actually add that because it just makes it much easier because, you know, now I know that, okay, this particular job uh matches my resume. um you know has a good matching criteria and so basically this there are more chances for me to actually get this job. Uh so I'm going to apply to it rather than apply to everything. So let me just delete this so that we have other data as well. So I'm just going to remove that. Okay. Now let's go back to this connect this and this is connected. So I'm going to save

it. All right. So now let's go through this particular one. All right. So now this is tools. What is this? This is nothing but a variable. So I've gone to tools and when you go to tools you have something called set a variable. All right. So I've used this particular one and what I essentially want to do is I want to put my resume in. All right. So for the matching criteria I want there has to be you need to give a chart GP how is it going to match right? So it is going to

take your resume and it's going to match with the job description which is extracted from the RSS feed and then it'll generate a score. So now this particular I've generated a text. So if you see I just went to LinkedIn I went to my profile and uh you know and out here basically. So what I did is I just copied everything from here. So I just open that and I copied everything from here and into a text document and uh you know this also as well as uh about and everything and then I create a

text document. Yeah. So I created a text document called uh resume text version and I've created this basically that's it. Okay. So this is all copy paste. So what I did is I just copied this from here. That's it. And then I went and paste it right here. That's it. Okay. So what's going to happen is that I'm I'm creating a variable. I don't have to create a variable. But then again I like doing things in a certain way so that it's not very messy and I can go and change it uh later on. If

I want to, you know, make some modifications to my resume, I can do that. And then I don't have to go into chat GPD modules to change it. I can just come to this particular variable, change it out here, save it, and it'll work perfectly fine. Okay, so that is the reason you use a variable and um one second just let's straighten these out. Okay, perfect. So now I have the resume, okay, within the variable. So now after this what I'm going to do is I'm going to add another chat JPD module and within the

module uh this is to transform. So this there's the format uh this is the prompt which I'm writing format this description properly so it text can be added to Google sheet without any formatting issues. Okay. So all right what happens is so this is particular again transform into structured data. So there are two you have different different modules for uh chart GPD right so when you go to chart GPD open it you have something called transform into structured data or create a completion prompt then all those things so these two we'll be using so the

reason I'm using the first one as transform is because when I get uh the description from let me just show this to you okay when the description comes in it comes with all the HTML tags so let me open this thing so if you can see with this particular chat GPT module what it does is it actually just cleans up that so this is what it's doing so this particular uh particular module what it do is it will take the description it'll remove any tags from there and uh after that it will just clean it

it's just format it cleanly so format the description properly so you uh so the text is can be added to Google sheet directly okay and it show me this paragraph so this is the prompt which I use and then of course I have created the structured data so I've u just added This parameter is description, job description, text. That's all. All right. So now what it's going to do is it's going to take that description from the feed, clean it up, and then add it to the sheet. Okay. So now this particular portion is done.

The next one is the completion report. So what this particular portion is doing is all right. So the first we have got this. Now the next one is when you're doing this, we have a summary right here. This is a job summary because the description becomes very long. Okay. Okay. So that is the reason I wanted to generate a summary and for that I'm using another module right here and the prompt for this is that take the job description which you're getting and then generate a summary for that you know so this particular so it

takes description which is coming from here and then I want you to read this LinkedIn job posting data and only output a job summary with a maximum 20 uh 215 characters. the job need must be I mean so I've written a prompt here basically which actually does is it kind of reduces it uh so it's more manageable all right so if you see out here if you go here you can see the summary the Commonwealth Bank this is so let me just show you the description as well I've hidden that so this is the description

okay this is all this this is full description but let me just hide that now Okay. And what I've done is I've just used that to generate the disc uh the summary so it's easier to read. Okay. And uh then after that we generate the uh the summary and after that again this is the third module which you're using the create a completion. This is where it actually I tell it that okay use this particular not the summary but the description which we're getting and the resume and then generate. I don't want to use a

summary for uh generating because it can uh for the generating the matching score because it is quite possible that you know it is um it misses some important points within the summary. So that is the reason I want to use the actual description to generate the score. Okay. So based on the job description which is coming from here where is the job description? Okay. Which is coming from the first one. This is where we actually get the description from. Right? So I'm saying that okay from here uh use this description and then generate a matching

score and use the uh the rum as well. So this is the based on the job description and the candidates resume. Your task is to analyze both and determine how well the rum matches the job description and then I've written a prompt which actually you know just outputs the number and uh it there should be a rational as well. So I want to know exactly what the chart GP's rational is. Why did it give 70? So it needs to explain that as well. Okay. So you know exactly it is just not making numbers out of

thin air. It has a logic behind that. And then here's the candidates resume. And as you said as I said uh we have a resume in the text variable. So we just use that. Okay. That's it. So this is where the resume is coming from right here. So we just pick this up and you put it here. Okay. All right. So this is the third module. And then the fourth one is this is done. And then after that again you transform the structured data. Now I uh what am I transforming here? Okay we are getting

two values from chart GP in the previous version right. So we are getting a rational and we getting a score and then I want to split it up again into two different um so then it comes two different columns. So then the match score comes like this and the rational comes like this right for that particular score. Okay. So that is the reason I've used the fourth module. And the finally I'm going to use again Google sheet and update a row. So this is the row which we have done. Uh I'm going to update the

remaining values. So again you go to drive spreadsheets uh sheet name row numbers. This is important and this is what you get from the first one. Okay. So you need to know the row number which you have right here. Uh one second. Google sheet search rows. You have a row number right here. So you will use this row number so that the correct row is updated. Okay. And that's all. So once you have that, you can save that and you can run it. So that is the whole scenario. So let me just take you through

this one more time end to end so that you know exactly what it is. All right. So the first thing what you need to do is you will have a timer and the timer actually tells you exactly how many minutes you want. Okay. So in my case I've used 4 hours but you can use whatever you want. So then after that we use a JSON. JSON basically what it does is you can uh you know it needs uh we need to actually loop through all the URLs rather having six different scenarios each with its own

URL we'll use the same scenario but we'll loop through all the URLs so if you want let's say you you're looking for jobs and there are 10 URLs you know maybe you're looking for in Brisbane you're looking for Adelaide as well or you're looking for Perth as well so you can just keep on adding to this URLs and what it's going to do is it's going to cycle through all of those okay so this is the iterator which actually cycles through all the URLs one by one. Then it goes to the RSS feed. We have

already created the feed. So it basically this is what scrapes the data and gets the data in. Then this is the module which we use uh to search if this particular URL exists there or not. If the jobs URL already exists like this then it will not add it. It will filter out. Okay. Right here this is the filter. Okay. After that the next step is uh what is that? transform into structured data. Again, I forgot this one. What was let me just uh remember it just quickly. Okay, so break. Okay, so yeah, so this

is what it's doing. So we are getting the job title uh these three fields as just one uh block and we are breaking it down so it's much more easy to read. So into job title, company and location. Okay, so this is what this module does. Then after that you have a set variable. This is to format the date. Then after that so we have all of those uh we have the date as well. So we directly add it to Google sheet. So in this case this particular module we'll have five uh so we add

the job title we add the company location post date and job link. The rest of the ones we'll update later. All right. So after that again I come here then I go to set variable. I've put in my resume here. Now this particular one uh what it does is it takes the URL which is coming from u uh the RSS feed and it cleans it up and it adds the description. Okay. So this is what I need a description is coming from here. Then it goes here in this module. What we'll do is again it

is create a comp uh completion. So again let me just see what it is doing. Again I forgot. Okay. All right. So what it does is in this job I want to okay so this is generating the summary from the description. So this gets the description this is fed into this one this particular generator summary which you see right here. Okay after that the third module is it actually uh data completion again this generates the score. So what this particular module does is it actually takes my resume and it takes uh the result which is

coming the job description and the resume. This is the job description and this is the rum and the prompt basically compares and generates a score and a rational. Okay. So we are again getting score and rational in just one field. And this fourth one what it does is it breaks it down into structured data. It will break it down in such a way so that we can get match score as well as rational at two different fields. Uh okay. And uh that's it. And then after that the last module we will update that. Okay. So

now I'm going to show you and we're going to run this whole thing. And let's uh of course we can't run for a long time because otherwise it'll just keep on running. There are six URLs each with maybe about 25 30 jobs. So you're talking about 160. But I just want to run it and show exactly how it's going to work. Okay. So let me just remove everything from here so there is no problem. And uh I'm just going to delete the rows from here. Uh-oh. Okay. So, I'm going to delete these rows. Okay. So,

that's now it's totally clean. There's nothing out here. Let's say. All right. And what I'm going to do is I'm going to run this now. And you can see it's running one. It goes here, goes here, here, here, and it's passed through because there's nothing. So, it'll add all these things. And now it is generating the description. Now it is generating the score, cleaning up the score right here. And then it updates the record. There you go. And then it is basically matched 50%. Okay. Why is that? Because it's UiPath developer. I'm not developer. So

it has generated and that is the rational. If you can see the rashan has extensive sequence the candidate seep kasta has extensive experience in digital transformation AI automation and SAS entrepreneurship but then however the specific requirements mention the job description for UiPath RPA developers such as practical expertise in UiPath uh power automate API so all of these things it requires all of these things in the job description which I probably don't have and that is the reason it is generated 50 score product owner rosewood Melbourne And then again I've got a score of 40. So

it will probably give you that rational as well. Uh the rationale is that however this role product owner specification requires experience in medical device section. Okay. So yeah I don't have that. So that's the reason. Okay. I've got a 50 again. So I've put in a filter. I mean like so much easier. What I've done is I've done gone to format and went to conditional formatting and I've put in that anything above 75 just give me a green. So it's easier to recognize. All right. And that's it. So that's what it's doing. And this keeps

on going right now. Okay. And that's it. It's pretty straightforward. So let me just run this and then show it to you how it looks. Okay. Okay. Let me just go back now. Yeah. Another thing I want to show you basically one is that. So now let me just stop this. Okay. four. Stop it. Requesting stop. Okay. So, now it's stopped. So, let me just delete that now. One. All right. So, we already have four of these, right? And we don't want duplicates. Now, when I run again, the same jobs are going to come and

but it should get filtered. That's what we need to see. Okay. And it should start with the fifth job. So, now let me just show this to you. What you'll see is like the filter this particular filter. Okay. If you see this, it is all green, but this should be red. Basically, should not let every URL pass through. All right. So, let's run this again. Again. Now, let's see. 2 3. See, right here, it's it's not letting it go. So, this the fifth job has gone through now. And that is the reason it doesn't come

here. So, you can see the fifth job. Okay. All right. Perfect. So, now this is working fine. And um let me just run this whole thing and I'm going to show you the result. Okay, great. All right, so now it's finished running and uh as you can see there is uh you know it's it's extracted about wow 148 jobs. Okay. And uh we have a scoring as well. So okay. So let's go down. Let's see. Um okay this is 90. This is interesting. So what is this? Principal product manager Gen Ei agents by for a

company called deputy and then it's in Sydney. I'm interested in this roles. Let me see this role as well. Okay. So this is the one to apply. Okay. And what I need to do is let me just go back and see the rationale behind it. Now this is the job summary which is there for the deputy. You can read that really quick so you know exactly whether it matches or not. And then I've got 90. So why is that why does it think like that? So let's see the candidate Sandep Kaista possesses a robust background

that aligns with the job description uh for strate strategic product leader with 20 years of experience in AI. This is there SEP exceeds um sorry I can barely read that SEP exceeds the required seven plus years of experience his roles at flipes and shuff demonstrate the so basically it's given you a reason as to why it thinks like that. So you know I have SAS I have so whatever there I have AI automation I have SAS in my resume and that is the reason it is giving you this particular score and where's that score okay

there you go okay so that's pretty much it here and so it's easier for you to actually see um and the benefit of this platform is that you don't have to go again and again into LinkedIn you you probably won't miss anything you just leave it and this keeps on actually updating automatically and if you don't like something you can delete that Or you just can hide it rather than deleting it because otherwise it'll add the row and uh we can really build on this. This is just a rudimentary system uh you know at least

the first one part of it. Okay. And so that there you go. This is exactly uh how this whole system works. So now if you want to see more uh into this automation, you know, this is a basic automation. If you want to actually add more filters to this thing, you can do that uh to make it more sort of customized to you. But if you would like to see more uh more in this automation something like let's say the system you should automatically scrape the website and then and generate uh the resume based on

the job description or some other idea feel free to leave that in the comments uh and uh I'll see what I can do if in case it's something which is very very interesting a lot of people actually uh want that then we will modify this and I'm going to create a part two of this where we use u the scraping results and then you know see the job description and your resume and kind of generate something automatically so that you can directly apply and it is like totally customized to that particular job role you know

and that's a wrap if you need help creating a voice AI agent or have any other AI automation project in mind feel free to book a consultation call with us links in the description make sure to check out our voice AI playlist for more in-depth use cases or click on another video I'll just picked for you thanks for watching I can't wait to see you in the next