← Video Library · AI Tools
The AI That Writes Your Cover Letters
Published November 22, 2024 · 37 views on YouTube
Key takeaways
- Turn your LinkedIn profile into a detailed PDF 'CV' with skills spelled out for each role, then let AI compare it against a job description.
- Ask the AI to score your fit on a scale of 1 to 100 before applying, so you don't waste time on jobs you don't qualify for.
- Have the AI identify the skills gap between your CV and the job description, then explain your transferable skills in the cover letter to close that gap.
- Set explicit guardrails so the AI only uses information you provided and never invents experience you don't have.
How can AI help you write a resume and cover letter that pass ATS screening?
By comparing a detailed PDF built from your LinkedIn profile against a job description, an AI tool can score your fit, identify missing skills, and rewrite your resume and cover letter with the keywords and phrasing an applicant tracking system (ATS) is looking for, without inventing experience you don’t have.
What is the workflow for turning a LinkedIn profile into a job-matched resume?
The system described in this video, called Profile Pilot, follows a repeatable sequence: build a detailed LinkedIn profile, export it as a PDF, then feed that PDF and a target job description into a custom ChatGPT setup.
- Build out your LinkedIn profile like a full CV, writing out what each company does, what you did in that role, and the specific skill sets tied to that job, rather than the shorter phrasing typical of a standard LinkedIn profile.
- Export the profile to PDF from LinkedIn’s Resources tab (save to PDF).
- Copy the job description, including qualifications and preferred requirements, from a job posting.
- Upload the profile PDF and paste the job description into the custom GPT.
- Let the AI analyze the match first and return a score out of 100.
- If the score is high enough, have the AI generate a new resume drawn from the CV that targets that specific job.
- Generate a cover letter that addresses the skills gap by explaining transferable skills for whatever percentage is missing.
- Ask the AI to re-score the new resume and cover letter against the job description to confirm the improvement.
How does the AI scoring system decide whether to apply for a job?
The tool rates a candidate’s CV against a job description on a scale of 1 to 100. In the video’s demonstration, a job scored 68 out of 100 before edits; after the AI generated a tailored resume and cover letter using transferable skills to cover the gap, the same job scored 85. The presenter’s rule of thumb: pursue a job scoring over 80, consider one around 70, and expect roughly a 10-point increase after the resume and cover letter are rewritten.
Does the AI make up experience or skills you don’t actually have?
No. The presenter built the system with a guardrail that prevents it from pulling information from the internet or inventing experience. It only draws from the profile/CV the user provides, and can prompt the user with questions (about volunteer work, informal experience, or other overlooked activities) to help them articulate skills they already have rather than fabricating new ones. This matters for the interview stage, since a candidate needs to be able to speak to anything written in the application.
Can this approach work for a career change or for someone without a detailed resume?
The presenter described using the same method for career changes: the AI reviews existing experience and reframes transferable skills for a new industry. For someone without a detailed CV, including high school students, the AI can ask questions about volunteer work, informal jobs (like babysitting or a newspaper route), and other unrecognized experience to help build out a usable skills profile from scratch, then provide a fill-in-the-blank template for a profile summary rather than writing it outright.
Full video transcript
Hello everybody and welcome to the Scale with AI Summit. My name is Kelly Tenny, one of the co-hosts, and we are diving into one of our hackathon mini sessions. I’m super excited these hackathons are going to be amazing because they are going to be short, quick, easy, implementable ideas that are really going to help you scale your workflows, your business, and all of the things that typically bog you down. And today we are here with the amazing Tyrese Niswander. Welcome, Tyrese.
Hi, how are you? I am great, and I’m so happy to have you on, because Tyrese, I know when I think of what amazing females do I know in my space that literally can create any AI hack to make things faster, more efficient, you are literally the top of mind. So I’m super happy to have you on.
You know me, I don’t like to be inefficient at anything, and it takes too long, I don’t have time for that.
I know, I know, you are like top five hackathon queens as far as I’m concerned. And I know that you’re here to show us something really cool that actually sprouted from your own personal need, which I love.
That’s right, that’s right. Just to get a little background from where I’m coming from is that I used to, first of all, operations is my jam, and so I always try to make things more efficient. So when AI first came out, I was really super interested in it. In fact, I got into Jasper AI before it was Jasper, it was Conversion AI I think, the very first iteration, which was way back in 2020. And so from 2020 all the way through to now I’ve moved on to different things, and now I’m actually a chief AI officer, as you can see my name there, CAIO is for chief AI officer, I can actually say that. Yeah, I actually know what I’m doing, and the really cool thing has been that with every iteration of ChatGPT and other tools I’m finding out it’s not about the tools, it’s about thinking about how you’re going to use the tools. Because unless you really are a coder or someone who already has that flow work mindset, you’re not going to understand how to actually ask the question in the prompt in the first place, and people say, oh, just talk to my chat like an assistant, no, but not really.
So anyway, bringing me to my app that I’m going to be talking about today: at the beginning of the year, in February, I was laid off because the company I was working for got actually bought out by a bigger company, and they didn’t need any marketing people. And I thought, okay, maybe this is an opportunity for me to jump into the chief AI officer certification, and that’s why I looked at that as an opportunity and not a setback. But in the meantime, to get unemployment I had to apply for three jobs a week, and I went to the Washington State work source office, and then I started to figure out, oh my God, they’re using computers, it is so complicated now to apply for a job, it’s beyond ridiculous. The ATS systems, which is the applicant tracking systems, look for keywords, they look for skills, soft skills, hard skills, emotional intelligence, all this stuff is all in there. And I thought to myself, the first resume that I did myself by hand, like we did in the old days, the old-fashioned way, and it took me hours upon hours. I was like, I can’t get three done in a week, what am I gonna do? And so then I thought, hold on, I’m a chief AI officer, I have skills I can bring to bear, so off we go.
I then went into ChatGPT to then go ahead with really building it out, and so I’m going to just share with you really quickly. My, can we go there, there we go, can we see my screen? It is a little delayed, there we go, okay. Basically, what I did was I created this because I really wanted to show that the traditional way was the long, hard way, and it was not something that everybody should be going through. I really feel like I want to save everyone from that experience, and so that is why I came out with this idea: that if you want to create a resume, first of all, the resumes you create, you don’t even know if you actually qualify for that job or not, because when we look at a job description we think to ourselves, oh yeah, I can do that, or I can do this, but can we translate that skill set into the resume so that it actually gets us the job, that’s an entirely different matter.
So what I did was, I got my LinkedIn profile, and my LinkedIn profile is built out completely like a CV, so it has all my volunteer work, all my experiences that I’ve had over my entire 57 years, if you will. And then I actually take that, make it into a PDF, then I ask the AI to compare my CV to the job at hand. Now here’s where it gets interesting: I’m telling it first to rate it on a scale of 1 to 100, so that it knows what my score actually is, so I’m not wasting time applying for a job I actually don’t qualify for. And then if it’s over 80 I go for it, if it’s around 70 or so I think, maybe I’ll just play around with it to see how it goes. And at the same time it’s going to look at the skills gap, that is all the skills that are missing, and then keep that in mind for my cover letter, which I’ll explain in a second. So the next thing that happens is that it creates the resume, now you have a new resume from that CV that goes with that job, which will pass the ATS system requirements.
Now the cover letter, what I, when I ask it to do that, it’s going to take that, let’s say it was 80% and there’s 20% missing, my applications is going to go back to the 20% that’s missing and say, hey Tyrese, you can use these different skills in its place, and the transferable skills are then explained in the cover letter to then increase your score. And then what we end up doing is we actually end up being able to create an application that is robust, that will actually get you the job. And better than that, the score that you started with will increase by the end. And at the end you can also ask the application, hey, the new resume and cover letter I just created, how well does it suit now to the job, and most of the time you get a 10-point increase, which is you can go from 80 to 90, 70 to 80. So then you have a better chance of getting that job. And so with that I can actually do three resumes and cover letters and applications in one hour, none of this days and days like I did before, it was all in one hour, I can do three in one hour. And I told my caseworker that, she says, so you’re getting paid $672 an hour, because that was what I was getting paid a week for my unemployment. So this is why I thought I can’t keep this to myself, I need to share it, because there’s so much here to understand, and really you’re not only just going to get the job, but you’re going to get a better job at a higher rate of pay, and sooner.
And if you’re doing a career change, it’s the same thing. If you just are going to take the application, it’s going to take the transferable skills and everything else that you have from your industry that you’re in and see how you can put it into a new industry so that you can make that change, because obviously if you’re changing industries you may or may not have the experience, but this way the application will actually help you to do that. And so once again, it’s all about the efficiency and cost effectiveness of life, and that’s where I live. And so that’s why I feel like my app, which is called the Profile Pilot, is the thing that I want every person to have who’s unemployed. And so with that I’m going to, oh yeah, if you guys want access to this thing I’m going to put a wait list going, and I’ll put a description in the link where you guys can click on that, and then you can get in there.
Okay, and now let’s go over to LinkedIn. Okay, so on LinkedIn, now what used to happen is, right in this area there used to be three little dots where you could click, and then it allowed you to optimize and change the resume according to the job, so you could actually create it manually. Now you can’t do that, they want you to pay for this. So here’s where you can go: once you have optimized your LinkedIn profile, and you’ve got all the experiences and everything that you want in here, then you can go to the Resources tab, and then you can click save to PDF. Once you save to PDF, that is your golden nugget. And from here, now what we’re going to do is we’re going to look at jobs. Let’s pick something here that’s already here, let’s try, see this looks interesting, Best Energy USA, okay, let’s try this one. Now all we’re going to do is we’re going to take the “about the job,” I haven’t even read it, right, I’m just looking at, wow, 75 to 100,000k, let’s see what happens. Okay, so you’re going to take all the job description and the qualifications that are required, and the preferred, things that we offer, doesn’t matter, and we’re going to copy this, and we’re going to head over to ChatGPT.
ChatGPT, we’re going to put this into my Profile Pilot, and we’re going to add my profile PDF if it decides to come up. Profile PDF, let me, is that your resume? That’s right, that’s the PDF that I just created out of the resources tab in the LinkedIn profile. Gotcha, okay. So now we’re gonna get that to get in here, and now it will, it knows what to do. So as soon as you do that, it’s going to ask, do the analysis, that’s the first thing that’s going to happen. So here it says 68 out of 100, so obviously I don’t qualify for this job, but let’s see, let’s follow this through and see what happens if I don’t touch anything and just say, okay, go ahead. It’s going to create a resume based on my CV and my LinkedIn profile, and it’s coming up with the whole thing. So you can basically copy and paste it into a Word document, create a PDF, and off it goes to whoever it is you’re going to apply to. But you want to make sure, as you’re reading through, to make sure that the details are correct, and nothing has been, I have programmed it so nothing is being pulled from the internet, it’s only pulling from your particular profile, because you don’t want to over-embellish or say things that you can’t do, because at the end of the day it’s too costly for a company to hire people, and so that’s why I took that into consideration.
Now let’s move over to the cover letter, let’s say okay fine, let’s go with the cover letter. Now it’s creating the cover letter. Now if I wanted to I could attach an example of a paragraph of a text that I wrote so that it could be in my writing, if I wanted to, it already knows my writing well enough here. But so here’s what the last thing we can do here, let’s see, so we started off at 68 out of 100, let’s see what is my new score now if we compare the new resume and cover letter to the job description. It jumped to 85, because it took all the information and took all transferable skills and everything, so it took that information, and now you’re basically able to apply for this job because now my resume is showing up at 85% connection rate. And that is why, and we’ve done that, what, we’ve only been on what, 10 minutes or 15 minutes, and we’ve done it already.
Yeah, pretty amazing, I’m like speechless, and I have some questions. Okay, fire away. So right, so originally you just, you copy that job description, right, and you were like, okay this maybe is one that LinkedIn thinks I’m going to be interested in, you pulled your resume, it gave you the score. So tell me a little bit more, the score is based on what, skills the resume has, and it’s comparing it to the skills that the job description is asking for, is that correct?
I should back up a little bit, that LinkedIn profile of mine, it has what I’ve done with each individual job is that I have actually gone away from your traditional LinkedIn profile, what I’ve actually written out is what the company does, what I did in that company as a job, and then added all the different skill sets that are required to go with that job, so that there’s additional information that is now in that CV, because now you’ve got your whole life experience in that PDF that you’re trying to apply with that job. So it’s not just that you’re taking your LinkedIn profile, but you’re taking the time to do all that one time, and then you can now apply for anything that you want, and every time that you want to update LinkedIn with a new job, you just do the same thing, and so then the way to apply would be seconds.
Okay, so that is the pre-work, right, because my brain is like, so are you telling me if I just give it my current CV and hook it up with a job, I’m imagining it will still give me a score, but it’s going to be way more accurate if I take the time to actually sit down and go through and add all of the skills and all of the components that were required of me. I do have information about that too, I have prompts that I built.
Ah, sure you do. Okay, so then my second question is, and we all just saw this workflow that you walked us through, so thank you for that, but so I noticed that it rewrote, right, like you asked it to rewrite your resume to actually, that’s the main difference is that it was a CV, and CV is a really long document with everything you’ve ever done in your life, the resume is simply a snapshot of bits and pieces of that CV, that’s why we can say that it’s a new resume, and the original document we can call that an actual CV, curriculum vitae is the right word, I should say.
Yes, so what it’s programmed to do is to pull the information from your CV, and talk to me about, so does it compare it to the things that the job description is asking for, and then it writes in a language that’s going to be more of a match for the wording that the job description is asking for, is correct? 100% yep.
Okay, so all the keywords and all the different skills that they’ve written in there, all of that is looked at. I actually took a class on ATS systems to understand it, because I thought, oh my, and it was a three-hour class, can you imagine, so I had written copious notes, and then I took all of that and I put that into Claude, and I asked Claude, okay, I need prompts out of this that will then be used for an application, can you help me with that, and that’s how I ended up doing that.
Okay, that’s amazing, and I like that you were able, it’s able to actually again give you this updated analysis, so going back and seeing the comparison. I also love that you’ve been able to apply it to, what if you’re looking for a new job in a new industry, so tell me, if someone didn’t have a CV, or if someone’s CV wasn’t as detailed as it needed, as it should be, to create for this prompt, how would we know?
We could ask the AI to ask us if we’ve got any volunteer experience or any other things that we’ve done that we don’t recognize as job experience. There’s so many people that when they were kids they were doing newspaper routes, or they were doing babysitting, but these are all things that involve accounting and marketing and actually operations, and learning how to do the route in the right way, and that there’s things that can be brought to bear that we actually never thought of, because we thought, oh it’s just babysitting, or it’s just a newspaper route, but these are things that high school students can actually bring to bear, because that’s another thing that I’ve been doing lately too, is that I’ve been teaching high school students how to use this particular application to then get their first job, and that is we’re using all kinds of volunteer experience, and then anything that they’ve done for somebody else just in a helpful way, they can use that to create the skill set that they need.
So can we practice, can you ask it for example to ask you some questions where the answers might help?
Ask me some questions basically to draw out information that might help with my CV, okay, I don’t have any experience, I want you to help me to, so here it’s now, I’ve got to move my little box here so I can read. I was like, okay so profile, now it’s referring to all the different sections that would be in your LinkedIn, meaning your transferable skills, relevant activities, education, informal volunteer experiences, the type of role that you’re interested in, and then it’s asking you to give a short introduction about yourself. And then let’s say, um, I don’t know how to write that short introduction, okay so can you help me with, so here we go, so you have your profile summary, now you’re just going to fill in the blanks, and then from the fill in the blanks now you can actually build on top of that and tweak it to the direction that you want it to go.
Amazing, and I love that it gave you a template, like it didn’t just say here’s something you could write, I see how this template is so valuable, just moving forward, for someone especially like you mentioned a high school student, someone coming out of college who was really, doesn’t know how to do it, I really like that, and I’m assuming you programmed it this way intentionally to be like, give me a usable template versus just spit out something on your own, so you’re teaching them, like you’re giving them a tool, and you’re almost teaching them the skill by providing that template instead of having the AI completely do it for them.
You’re gonna think I’m crazy, but my prompt, it’s 11 pages long. Then it was a culmination of, oh I need to add this, no I need to add that, and then you just keep adding, and then the next thing you know you got this big monstrosity that you’re trying to bundle into this tiny app. That’s, yeah, so I have tried really, and the other thing that I really want to, here, which I express to the high school students, is that at no time do I allow the GPT to go out and get the information, it’s got to be your experience, your skill set, it does not add anything that you have not provided, it will give you direction, but at no time is it allowed to go beyond the guardrails that I’ve put on it, and that is that everything has to be done in an ethical and responsible way, because that is the only way I think AI should go forward in the first place.
Yeah, all right everybody, thank you so much for sharing this with us, and again, I hope everybody recognizes, right, and so thank you for this conversation, because one of the things I think this reminds us is, like you said, we, AI could make up a resume, AI could write my cover letter, I could, and I know that this is what people are doing, is they’re taking a job description and they’re putting it into AI and they’re saying create my resume so that I’m a good fit for this job and write my cover letter. And I love that you’re recognizing, obviously here, anybody at the Scale with AI inside Gen AI University, inside Women Who AI, we’re always about ethical uses, we’re always about authenticity, right, we’re always about making sure that we’re using AI as a tool and not to cheat, not as a replacement. And I love that you’ve set those guardrails up inside of your ChatGPT, and I think that is maybe even a great reminder for a lot of people to recognize, we can give it parameters.
100%, and I don’t know how many people are doing that, and I think that might be a takeaway for a lot of people just watching this hackathon session, is like, oh yes, I built GPTs, but I don’t give it parameters to be able to say only rely on the content that’s being uploaded, please ignore any content that is out there on the internet. Because at the end of the day, you know, if you get that interview and all of a sudden they’re asking you about an experience that you put on paper and you can’t speak to it, how embarrassing is that. And that’s the same goes for all those people that say, oh let me write a book for you, like no, if I can’t give you what is necessary to put in the book, how are you gonna write the book for me? How will I know if someone decides, oh Tyrese, I want you to come and talk on my webinar about your book, how are you going to address that? These things have to come to be, you have to be responsible, because at the end of the day your word is everything, your reputation, everything, and if people think that you’re a cheater, a liar, or any of those negative connotations, why would they ever hire you.
Yeah, yeah, and that’s the thing, at the end of the day. All right everybody, there you have it, thank you so much Tyrese, and this mini hackathon session, as she shared, she’s so graciously offering anybody a link that’s going to be in this description to this session, so I don’t know, let’s see how many people together a really good package for everyone so that they can use it, and then they can share it with their friends if they need it.
Yeah, let’s see how many people go and change jobs and change career fields, right, anybody who’s unhappy, I’m getting this link and I’m off to becoming a whatever my other dream, I’m gonna become a marine biologist, I think.
You knew that at some point I wanted to. That’s, I don’t know, marketing can go into marine biology just like that, it’s okay, I can’t be on a boat anyways, which is exactly why I threw that out the window a long time ago, but there you have it. So brilliant, such a great hack, on, thank you so much Tyrese for joining us in the Scale with AI Summit.
Thank you so much, I really enjoyed being here, and thank you for asking me to be here to talk to the community.
Yeah, yeah, take care everybody, we’ll see you in the next session, okay, bye bye.