WEBVTT 00:00:00.000 --> 00:00:02.180 Imagine a world where you never have to scroll 00:00:02.180 --> 00:00:05.419 a job board again. Seriously. What if a quiet, 00:00:05.540 --> 00:00:08.759 tireless AI could find your perfect role, customize 00:00:08.759 --> 00:00:12.539 your resume, and even draft the email? All of 00:00:12.539 --> 00:00:14.660 it. Automatically. You're talking about, like, 00:00:14.679 --> 00:00:17.239 basically having a super smart robot clone that 00:00:17.239 --> 00:00:20.140 just handles all the paperwork? Yeah. Yeah, that's 00:00:20.140 --> 00:00:21.780 pretty much the dream, isn't it? Welcome to the 00:00:21.780 --> 00:00:24.280 Deep Dive. Today, we're stepping into a fascinating 00:00:24.280 --> 00:00:26.820 new set of sources that promises to fundamentally 00:00:26.820 --> 00:00:29.820 shift one of the most, well... Let's be honest, 00:00:29.980 --> 00:00:32.520 soul -crushing tasks out there. Job hunting. 00:00:32.759 --> 00:00:35.179 Yeah, and we're diving into building your very 00:00:35.179 --> 00:00:37.640 own AI system for this. And here's the cool part. 00:00:37.799 --> 00:00:41.079 No complex coding needed. It's about pure ingenuity, 00:00:41.079 --> 00:00:43.679 really. Just leveraging smart tools. It truly 00:00:43.679 --> 00:00:46.479 is a game changer. Our mission today is to unpack 00:00:46.479 --> 00:00:49.179 a detailed step -by -step blueprint for constructing 00:00:49.179 --> 00:00:52.649 a personal AI job application system. We'll explore 00:00:52.649 --> 00:00:54.789 why the modern job search often feels like a 00:00:54.789 --> 00:00:57.770 full -time job in itself and how this AI can 00:00:57.770 --> 00:01:00.570 work tirelessly for you 24 -7. We'll trace its 00:01:00.570 --> 00:01:02.990 philosophy, the specific digital tools you'll 00:01:02.990 --> 00:01:04.709 need to gather, how it discovers and filters 00:01:04.709 --> 00:01:07.829 jobs with real precision, tailors your applications, 00:01:08.030 --> 00:01:10.450 and even handles that initial outreach. Plus, 00:01:10.629 --> 00:01:12.650 we'll touch on troubleshooting common hiccups 00:01:12.650 --> 00:01:14.810 and, of course, the responsible side of wielding 00:01:14.810 --> 00:01:17.069 such a superpower. Get ready to shift from being 00:01:17.069 --> 00:01:21.030 a reactive applicant to a truly proactive problem 00:01:21.030 --> 00:01:23.189 solver. Okay, let's first consider the universal 00:01:23.189 --> 00:01:25.670 pain points of job hunting. Our sources paint 00:01:25.670 --> 00:01:29.000 a really clear picture. The endless scrolling 00:01:29.000 --> 00:01:32.500 through just countless job boards. The tedious 00:01:32.500 --> 00:01:35.159 necessity of customizing your resume for every 00:01:35.159 --> 00:01:37.219 single role. Oh, yeah. Crafting those unique 00:01:37.219 --> 00:01:40.640 cover letters and then that often dreaded black 00:01:40.640 --> 00:01:43.180 hole of silence. No replies. It's repetitive, 00:01:43.500 --> 00:01:47.519 incredibly inefficient, and undeniably discouraging. 00:01:47.659 --> 00:01:50.459 Yeah. You often feel like just another data point, 00:01:50.579 --> 00:01:52.719 right? Just one of thousands in a digital cube. 00:01:52.959 --> 00:01:55.040 Yeah. It's exactly like throwing darts in the 00:01:55.040 --> 00:01:57.659 dark, just hoping one sticks. That whole cycle 00:01:57.659 --> 00:02:00.379 of hope and then disappointment really does wear 00:02:00.379 --> 00:02:02.900 you down. Who hasn't felt that immense pressure, 00:02:03.000 --> 00:02:05.340 that feeling of sending your carefully crafted 00:02:05.340 --> 00:02:08.360 application into what often feels like, you know, 00:02:08.379 --> 00:02:11.060 a bottomless digital void? It's tough, really 00:02:11.060 --> 00:02:13.139 tough. But here's where it gets really interesting. 00:02:13.960 --> 00:02:16.259 What if you could actually change the rules of 00:02:16.259 --> 00:02:18.819 that game? What if you didn't have to just accept 00:02:18.819 --> 00:02:21.659 that exhausting cycle? Our source proposes building 00:02:21.659 --> 00:02:24.520 a personal AI system that tirelessly searches, 00:02:24.699 --> 00:02:27.539 crafts personalized applications, and operates 00:02:27.539 --> 00:02:30.520 with a speed and precision that are simply, well... 00:02:30.840 --> 00:02:33.800 humanly impossible. It's truly tireless, isn't 00:02:33.800 --> 00:02:36.020 it? Completely. And it can do this for hundreds, 00:02:36.120 --> 00:02:39.460 maybe even thousands of jobs in the time it takes 00:02:39.460 --> 00:02:41.259 you to manually apply for just a handful. And 00:02:41.259 --> 00:02:43.840 the best part, it genuinely claims no coding 00:02:43.840 --> 00:02:45.919 knowledge is required. This isn't about diving 00:02:45.919 --> 00:02:48.919 into complex programming languages or like syntax. 00:02:49.419 --> 00:02:52.199 Instead, it uses simple visual automation tools, 00:02:52.280 --> 00:02:54.379 often with drag and drop interfaces. It's about 00:02:54.379 --> 00:02:57.020 building a consistent fatigue proof machine that 00:02:57.020 --> 00:03:00.099 truly plays the numbers game for you. Boiling 00:03:00.099 --> 00:03:02.539 it down, what's the core shift this AI system 00:03:02.539 --> 00:03:05.259 enables for job seekers? It flips the power dynamic, 00:03:05.460 --> 00:03:08.360 really. Moving from reactive applying to proactive 00:03:08.360 --> 00:03:12.199 automated deployment of your skills. Okay. Beyond 00:03:12.199 --> 00:03:14.740 just the raw speed, there's a powerful philosophy 00:03:14.740 --> 00:03:18.219 woven into the fabric of this system. Show, don't 00:03:18.219 --> 00:03:21.520 tell. It's about demonstrating your capabilities, 00:03:21.819 --> 00:03:24.699 not merely listing them on a resume. Yeah, this 00:03:24.699 --> 00:03:26.960 is the secret sauce, I think, especially for 00:03:26.960 --> 00:03:29.800 roles in tech, marketing, or operations where 00:03:29.800 --> 00:03:32.159 innovation and initiative are genuinely valued. 00:03:32.479 --> 00:03:34.979 Everyone claims to be innovative or tech -savvy 00:03:34.979 --> 00:03:36.879 on their resume, right? This system lets you 00:03:36.879 --> 00:03:38.840 prove it right from the very first interaction. 00:03:39.139 --> 00:03:41.400 Right. Imagine crafting an outreach email that 00:03:41.400 --> 00:03:44.120 doesn't just state, I'm skilled in AI and automation. 00:03:44.360 --> 00:03:47.039 Instead, it begins something like this. To be 00:03:47.039 --> 00:03:49.719 upfront, I actually used an AI system I built 00:03:49.719 --> 00:03:52.419 to find this job opening, customize my resume 00:03:52.419 --> 00:03:54.500 specifically for this role, and automatically 00:03:54.500 --> 00:03:57.639 gather your contact details. I believe in showing 00:03:57.639 --> 00:03:59.740 my skills rather than just talking about them, 00:03:59.800 --> 00:04:01.719 and I'd be happy to explain how it works. It's 00:04:01.719 --> 00:04:04.400 bold, it's direct, and it just arrives in their 00:04:04.400 --> 00:04:07.680 inbox. Wow. I mean, imagine being a hiring manager 00:04:07.680 --> 00:04:10.139 getting that email. It would land like a bombshell 00:04:10.139 --> 00:04:13.379 in an inbox full of generic applications. Yours 00:04:13.379 --> 00:04:15.319 immediately stands out. It demonstrates initiative, 00:04:15.680 --> 00:04:18.319 technical proficiency, and a results -oriented 00:04:18.319 --> 00:04:20.839 mindset right from the start. You're showing 00:04:20.839 --> 00:04:24.060 real -life work, not just talking about it. You've 00:04:24.060 --> 00:04:26.699 already provided a tangible example of your cleverness 00:04:26.699 --> 00:04:29.939 and capability. So how does this show -don't 00:04:29.939 --> 00:04:34.319 -tell approach fundamentally redefine a job application? 00:04:34.759 --> 00:04:37.019 It transforms your application from just a static 00:04:37.019 --> 00:04:39.839 document into a live demonstration of your actual 00:04:39.839 --> 00:04:42.800 capabilities. Okay. Every significant project, 00:04:43.160 --> 00:04:45.540 big or small, it starts with careful preparation. 00:04:46.199 --> 00:04:48.980 Think of a chef meticulously preparing ingredients 00:04:48.980 --> 00:04:52.160 before tackling a complex dish. For this AI system, 00:04:52.379 --> 00:04:54.600 our ingredients are a few key online services. 00:04:54.740 --> 00:04:56.699 And this initial setup, it's a one -time process. 00:04:56.899 --> 00:04:59.000 Once it's configured, it establishes the foundation 00:04:59.000 --> 00:05:01.279 for everything else that follows. First up, you've 00:05:01.279 --> 00:05:03.759 got NAN. This is your factory floor, basically. 00:05:04.250 --> 00:05:06.189 It's the visual automation platform where all 00:05:06.189 --> 00:05:08.730 your various digital tools connect. It's essentially 00:05:08.730 --> 00:05:11.290 a blank canvas where you connect different applications 00:05:11.290 --> 00:05:13.769 like building blocks, creating a logical flow 00:05:13.769 --> 00:05:15.949 for your information. Think of it, yeah, like 00:05:15.949 --> 00:05:18.649 stacking Lego blocks of data. Then we have the 00:05:18.649 --> 00:05:22.110 OpenAI API that acts as the brain of our operation. 00:05:22.779 --> 00:05:24.759 This provides the intelligence needed for tasks 00:05:24.759 --> 00:05:28.420 like filtering job descriptions and crafting 00:05:28.420 --> 00:05:31.100 personalized text. We also need Google Cloud 00:05:31.100 --> 00:05:33.279 Console, which functions as a sort of secure 00:05:33.279 --> 00:05:37.019 key master to allow N8n to connect safely to 00:05:37.019 --> 00:05:39.360 your Google Docs and Gmail, ensuring specific 00:05:39.360 --> 00:05:41.980 controlled permissions. Very important. Yeah, 00:05:42.100 --> 00:05:44.199 definitely. For scouting those job openings, 00:05:44.360 --> 00:05:47.120 there's Apify, our tireless scout for web scraping 00:05:47.120 --> 00:05:49.600 job listings. It automatically gathers the raw 00:05:49.600 --> 00:05:51.720 material, those job postings for your system. 00:05:52.170 --> 00:05:54.769 And any mail finder, which is like your private 00:05:54.769 --> 00:05:57.870 detective, to unearth direct contact emails, 00:05:58.129 --> 00:06:00.629 helping you bypass the general info at addresses. 00:06:01.149 --> 00:06:03.589 Finally, just a standard Gmail account for drafting 00:06:03.589 --> 00:06:05.930 those outreach emails. That's your communications 00:06:05.930 --> 00:06:08.310 hub, pretty much. Now let's talk about the cost. 00:06:08.550 --> 00:06:11.209 This powerful system, it isn't free to operate, 00:06:11.290 --> 00:06:13.170 of course, but the cost per application turns 00:06:13.170 --> 00:06:15.850 out to be incredibly low. It's best viewed as 00:06:15.850 --> 00:06:18.329 a strategic investment in your career rather 00:06:18.329 --> 00:06:20.649 than just another recurring expense. Our source 00:06:20.649 --> 00:06:24.009 estimates Apify at roughly, what, $1 per 1 ,000 00:06:24.009 --> 00:06:26.990 jobs scraped, OpenAI at approximately $0 .02 00:06:26.990 --> 00:06:29.790 per job processed, and AnyMailFinder at about 00:06:29.790 --> 00:06:32.370 $0 .03 per email lookup. That shakes out to about 00:06:32.370 --> 00:06:36.430 $5 .10 for 100 job applications. Think about 00:06:36.430 --> 00:06:38.449 that for a second. For the price of a decent 00:06:38.449 --> 00:06:44.399 coffee, you can process 100 applications. increased 00:06:44.399 --> 00:06:46.699 annual income from landing a successful job placement. 00:06:46.860 --> 00:06:49.899 The return on investment here is genuinely enormous. 00:06:50.160 --> 00:06:53.000 OK, a crucial component is your master resume. 00:06:53.220 --> 00:06:56.120 It needs to be simple, text friendly and stored 00:06:56.120 --> 00:06:58.899 in Google Docs. Crucially, it must use clear, 00:06:59.079 --> 00:07:01.740 semantically structured headings. Our source 00:07:01.740 --> 00:07:04.420 specifically stresses using built in Google Docs 00:07:04.420 --> 00:07:07.519 styles like heading two for major sections, maybe 00:07:07.519 --> 00:07:10.420 heading three for job titles. Why does simplicity 00:07:10.420 --> 00:07:13.660 matter so much here? Well, AI models primarily 00:07:13.660 --> 00:07:17.139 process text and structure. Complex visual layouts 00:07:17.139 --> 00:07:20.199 with multiple columns or unusual fonts can easily 00:07:20.199 --> 00:07:23.120 confuse them, leading to jumbled or incomplete 00:07:23.120 --> 00:07:26.000 outputs. Yeah, exactly. You really want to give 00:07:26.000 --> 00:07:28.160 the builder clean bricks, not a jumbled mess 00:07:28.160 --> 00:07:30.540 of construction debris. Setting up Google Cloud 00:07:30.540 --> 00:07:32.860 Console for those secure connections to Docs 00:07:32.860 --> 00:07:35.060 and Gmail, that can be the most time -consuming 00:07:35.060 --> 00:07:36.680 part initially, especially if you haven't really 00:07:36.680 --> 00:07:38.600 ventured into that kind of setup before. But 00:07:38.600 --> 00:07:41.829 once it's done, it's done forever. special specific 00:07:41.829 --> 00:07:44.810 keys, API keys, not giving away your main Google 00:07:44.810 --> 00:07:46.990 password, which is obviously a vital security 00:07:46.990 --> 00:07:49.810 step. So what's the fundamental principle for 00:07:49.810 --> 00:07:52.589 preparing your resume for this AI system? Simplicity 00:07:52.589 --> 00:07:54.850 and structured headings are absolutely key for 00:07:54.850 --> 00:07:57.889 the AI to process it correctly. No fancy stuff. 00:07:58.209 --> 00:08:01.350 Mid -roll sponsor read. All right. With our tools 00:08:01.350 --> 00:08:03.949 meticulously prepared, we now move to the first 00:08:03.949 --> 00:08:07.110 major stage of our digital factory, job discovery 00:08:07.110 --> 00:08:09.720 and filtering. This is where the system truly 00:08:09.720 --> 00:08:12.759 begins its tireless work. Okay, step one is making 00:08:12.759 --> 00:08:15.040 sure your master resume is read into the system. 00:08:15.180 --> 00:08:18.800 In N8n, you add a Google Docs node, provide the 00:08:18.800 --> 00:08:20.959 document ID from your resume's URL, just copy 00:08:20.959 --> 00:08:23.399 -paste it, and then you pin that output. This 00:08:23.399 --> 00:08:25.339 simply means the AI has your core information 00:08:25.339 --> 00:08:27.779 readily available for all subsequent steps without 00:08:27.779 --> 00:08:29.819 needing to reread it repeatedly. Pretty easy, 00:08:29.860 --> 00:08:32.000 right? Yeah, sounds straightforward. Next, the 00:08:32.000 --> 00:08:33.820 system begins scraping for new job openings. 00:08:34.320 --> 00:08:37.100 Now this step depends entirely on crafting the 00:08:37.100 --> 00:08:40.059 perfect search URL directly from LinkedIn jobs. 00:08:40.700 --> 00:08:43.440 Our source suggests using Boolean operators like 00:08:43.440 --> 00:08:48.379 project manager and TAC to precisely narrow your 00:08:48.379 --> 00:08:51.019 focus. You also apply filters like past -went 00:08:51.019 --> 00:08:53.279 and specific locations to ensure you're getting 00:08:53.279 --> 00:08:55.879 fresh, relevant listings. Once your LinkedIn 00:08:55.879 --> 00:08:58.019 search results look exactly right, you simply 00:08:58.019 --> 00:09:00.639 copy that entire URL from your browser's address 00:09:00.639 --> 00:09:03.279 bar. That becomes your direct input for Appify. 00:09:03.559 --> 00:09:06.600 Right. You'll then use an HTTP request node, 00:09:06.779 --> 00:09:08.559 which is essentially a digital messenger, to 00:09:08.559 --> 00:09:11.250 command Appify. specify that LinkedIn search 00:09:11.250 --> 00:09:13.590 URL and tell it how many jobs to scrape, maybe 00:09:13.590 --> 00:09:16.629 100 to start. Apify will return those 100 data 00:09:16.629 --> 00:09:18.590 outputs, which you can then send through one 00:09:18.590 --> 00:09:21.009 by one using something called a limit node. This 00:09:21.009 --> 00:09:23.129 is a neat trick to prevent a system traffic jam 00:09:23.129 --> 00:09:25.509 by managing the flow of data, making sure the 00:09:25.509 --> 00:09:27.990 process stays smooth and stable. Then comes the 00:09:27.990 --> 00:09:30.649 AI intelligence filtering. You add an open AI 00:09:30.649 --> 00:09:33.490 node, choosing a fast and cost -effective model 00:09:33.490 --> 00:09:36.870 like GPT 4 .1 Mini for this initial high -volume 00:09:36.870 --> 00:09:39.750 task. You then craft an about -me context paragraph 00:09:39.750 --> 00:09:42.659 for the AI. This goes beyond just listing skills. 00:09:42.840 --> 00:09:45.519 You should include career goals, preferred company 00:09:45.519 --> 00:09:48.779 culture, perhaps something like I thrive in collaborative, 00:09:48.860 --> 00:09:51.600 fast paced environments and even anti goals, 00:09:51.740 --> 00:09:54.399 which you explicitly want to avoid. Oh, interesting. 00:09:54.559 --> 00:09:56.919 Like what? Like roles primarily focused on individual 00:09:56.919 --> 00:09:58.980 sales quotas, maybe if that's not your thing. 00:09:59.059 --> 00:10:02.210 Gotcha. Yeah, this rich context truly helps the 00:10:02.210 --> 00:10:04.669 AI act as a personal agent, understanding your 00:10:04.669 --> 00:10:07.870 nuanced preferences. The AI's task here is actually 00:10:07.870 --> 00:10:10.929 quite simple. To return a straightforward JSON 00:10:10.929 --> 00:10:13.370 object, which is just a standard structured way 00:10:13.370 --> 00:10:15.809 of sending data, giving us a clear verdict, true 00:10:15.809 --> 00:10:18.350 or verdict, false, based on your fit for that 00:10:18.350 --> 00:10:21.490 job description. You then add a siltr or if node, 00:10:21.730 --> 00:10:23.830 allowing only the true jobs to pass through to 00:10:23.830 --> 00:10:26.070 the next stage. Another limit node here keeps 00:10:26.070 --> 00:10:28.269 things flowing smoothly, preventing any kind 00:10:28.269 --> 00:10:31.090 of overload. So how does the AI refine its job 00:10:31.090 --> 00:10:32.789 search to align with your personal preferences? 00:10:33.149 --> 00:10:36.070 It uses your detailed about me context, including 00:10:36.070 --> 00:10:39.110 those goals and anti -goals, to act as your truly 00:10:39.110 --> 00:10:42.309 personalized agent. Okay. For every job that 00:10:42.309 --> 00:10:44.990 successfully passes our filter, we move into 00:10:44.990 --> 00:10:48.210 the really crucial phase of application customization. 00:10:48.669 --> 00:10:50.769 This is where the magic really starts to happen, 00:10:50.929 --> 00:10:54.029 tailoring your approach to each specific opportunity. 00:10:54.330 --> 00:10:57.529 Exactly. So another open AI node, but this time 00:10:57.529 --> 00:10:59.730 you'll likely want to use a higher quality model 00:10:59.730 --> 00:11:02.929 like GPT 4 .1 for the actual writing tasks, as 00:11:02.929 --> 00:11:05.230 it requires more nuanced understanding and, well, 00:11:05.350 --> 00:11:07.909 creative output. You feed it your comprehensive 00:11:07.909 --> 00:11:10.070 about me summary, the specific job description 00:11:10.070 --> 00:11:12.389 for the role, and your master resume content. 00:11:12.649 --> 00:11:15.149 And you can guide the AI with explicit instructions. 00:11:15.590 --> 00:11:18.129 Tell it to mirror keywords found in the job descriptions 00:11:18.129 --> 00:11:20.590 requirement section, weaving them into your skills 00:11:20.590 --> 00:11:22.990 and experience sections. You also instruct it 00:11:22.990 --> 00:11:25.110 to maintain a professional tone and crucially 00:11:25.110 --> 00:11:27.769 respond only with the updated, customized resume 00:11:27.769 --> 00:11:30.590 and markdown format. No extra introductory text, 00:11:30.730 --> 00:11:33.629 no code blocks, just the resume itself. Right. 00:11:34.139 --> 00:11:37.000 Next, you convert that markdown to HTML. Why? 00:11:37.240 --> 00:11:39.480 Because document tools, especially something 00:11:39.480 --> 00:11:42.419 like the Google Docs API, they prefer HTML for 00:11:42.419 --> 00:11:44.840 preserving formatting. This ensures all your 00:11:44.840 --> 00:11:48.039 headings, bold text, and bullet points are perfectly 00:11:48.039 --> 00:11:51.830 maintained in the final document. Then a Google 00:11:51.830 --> 00:11:54.230 Docs node creates a brand new document dynamically 00:11:54.230 --> 00:11:57.070 named something like Tech Corp Jane Doe Custom 00:11:57.070 --> 00:12:00.049 Resume, you know, specific. An HTTP request node 00:12:00.049 --> 00:12:03.049 then patches this blank file with your HTML content. 00:12:03.250 --> 00:12:04.629 Think of it like giving a specific instruction 00:12:04.629 --> 00:12:07.259 to Google Drive saying, OK. Fill in the blanks 00:12:07.259 --> 00:12:08.860 of that newly created document with this content. 00:12:09.179 --> 00:12:11.399 And finally, a Google Drive node adjusts the 00:12:11.399 --> 00:12:13.659 sharing permissions, making the document publicly 00:12:13.659 --> 00:12:16.539 viewable via a link. This is the actual custom 00:12:16.539 --> 00:12:18.299 resume you'll share with potential employers. 00:12:18.799 --> 00:12:21.399 The final step involves finding the decision 00:12:21.399 --> 00:12:24.379 maker. You use another HTTP request node to connect 00:12:24.379 --> 00:12:26.740 to any mail finder providing the company's domain, 00:12:26.940 --> 00:12:29.340 like techcorp .com, and the type of person you're 00:12:29.340 --> 00:12:32.159 looking for, perhaps CEO or hiring manager or 00:12:32.159 --> 00:12:34.960 VP of engineering. A filter node then ensures 00:12:34.960 --> 00:12:36.919 you only proceed if a valid email address is 00:12:36.919 --> 00:12:39.220 actually found for that specific role. The very 00:12:39.220 --> 00:12:42.679 last step is crafting the perfect email using 00:12:42.679 --> 00:12:45.659 a Gmail node. The to address comes directly from 00:12:45.659 --> 00:12:48.240 any mail finder if it found one. The subject 00:12:48.240 --> 00:12:51.080 line is dynamic, for instance, regarding the 00:12:51.080 --> 00:12:53.639 project manager position to make it highly relevant 00:12:53.639 --> 00:12:56.220 and specific. And the body. It's that powerful. 00:12:56.840 --> 00:12:59.120 show don't tell message we discussed earlier 00:12:59.120 --> 00:13:01.360 something like hey decision makers first name 00:13:01.360 --> 00:13:03.480 i see you're hiring right now and i wanted to 00:13:03.480 --> 00:13:05.320 reach out to say that i believe i'm a strong 00:13:05.320 --> 00:13:07.899 fit for the role to be transparent i actually 00:13:07.899 --> 00:13:10.620 used an ai system i built to find this job opening 00:13:10.620 --> 00:13:13.320 tailor my resume specifically for it here's the 00:13:13.320 --> 00:13:15.340 link link and gather your contact information 00:13:15.340 --> 00:13:17.840 i believe in showing results over just talking 00:13:17.840 --> 00:13:20.159 so i'd be happy to walk you or any hiring managers 00:13:20.159 --> 00:13:22.440 through how the system works it includes that 00:13:22.440 --> 00:13:24.759 live link to your newly generated custom resume 00:13:24.759 --> 00:13:27.960 it's direct it's confident. How does this system 00:13:27.960 --> 00:13:30.840 ensure your customized resume looks polished 00:13:30.840 --> 00:13:33.240 and professional every time? It converts the 00:13:33.240 --> 00:13:36.399 AI's markdown output to HTML, which Google Docs 00:13:36.399 --> 00:13:38.559 uses to preserve perfect formatting, headings, 00:13:38.559 --> 00:13:40.820 and all. Okay, so building an automation system, 00:13:40.940 --> 00:13:44.220 especially one this comprehensive, it rarely 00:13:44.220 --> 00:13:46.559 works perfectly on the very first try. That's 00:13:46.559 --> 00:13:48.980 just reality. So let's talk about testing and 00:13:48.980 --> 00:13:50.759 troubleshooting. It's an absolutely essential 00:13:50.759 --> 00:13:53.980 part of building any robust system like this. 00:13:54.039 --> 00:13:56.740 Yeah, absolutely. A common issue you might encounter 00:13:56.740 --> 00:13:59.059 is an error like the resource you are requesting 00:13:59.059 --> 00:14:01.759 could not be found in ATEM. This often means 00:14:01.759 --> 00:14:03.860 your nodes are kind of out of sync, especially 00:14:03.860 --> 00:14:06.620 if you've rerun only parts of the workflow. The 00:14:06.620 --> 00:14:09.179 solution is usually to unpin all the data points, 00:14:09.240 --> 00:14:11.379 except for maybe very expensive ones like that 00:14:11.379 --> 00:14:14.120 initial Appify scrape, and simply rerun the entire 00:14:14.120 --> 00:14:16.360 workflow from the beginning. That usually resyncs 00:14:16.360 --> 00:14:19.080 everything. Another point to consider, low email 00:14:19.080 --> 00:14:21.950 discovery rate. Our source suggests that a 30 00:14:21.950 --> 00:14:24.590 -45 % success rate for finding direct contact 00:14:24.590 --> 00:14:27.289 emails is actually quite realistic. It's not 00:14:27.289 --> 00:14:29.669 going to be 100%. Larger companies often have 00:14:29.669 --> 00:14:31.990 more sophisticated protections and very small 00:14:31.990 --> 00:14:34.169 startups might not have much public data yet. 00:14:34.269 --> 00:14:36.990 The solution? Well, you might consider focusing 00:14:36.990 --> 00:14:39.210 your job search more on small to medium sized 00:14:39.210 --> 00:14:41.990 businesses. And importantly, ensure continue 00:14:41.990 --> 00:14:44.129 on error is turned on for that email finder step. 00:14:44.289 --> 00:14:46.070 So the whole workflow doesn't just stop dead 00:14:46.070 --> 00:14:48.529 if one email isn't found. Right. And, you know, 00:14:48.549 --> 00:14:50.529 I still wrestle with prompt drift myself sometimes. 00:14:50.570 --> 00:14:53.470 So here's a vulnerable admission. AI generated 00:14:53.470 --> 00:14:56.269 resumes can sometimes have weird formatting if 00:14:56.269 --> 00:14:58.850 the AI misinterprets your template. That's prompt 00:14:58.850 --> 00:15:01.590 drift when the AI subtly deviates from your initial 00:15:01.590 --> 00:15:04.250 instructions over time, often leading to unexpected. 00:15:04.679 --> 00:15:07.759 or slightly off outputs. This might be from complex 00:15:07.759 --> 00:15:10.840 layouts or like hidden tables in your original 00:15:10.840 --> 00:15:13.679 resume. The solution, simplify your master resume 00:15:13.679 --> 00:15:16.940 even further. Seriously, no hidden tables. Keep 00:15:16.940 --> 00:15:19.919 it truly clean. And explicitly tell the AI in 00:15:19.919 --> 00:15:22.539 your prompt, do not add any introductory or concluding 00:15:22.539 --> 00:15:24.899 text. Respond with only the resume content in 00:15:24.899 --> 00:15:27.299 markdown. Be very direct. What if the workflow 00:15:27.299 --> 00:15:29.740 feels sluggish or even times out completely? 00:15:30.440 --> 00:15:32.860 Processing 100 or more jobs all at once can simply 00:15:32.860 --> 00:15:35.580 overload the system. system, especially the free 00:15:35.580 --> 00:15:37.899 tiers of some services. The elegant solution 00:15:37.899 --> 00:15:40.000 here is to use that limit node again, but this 00:15:40.000 --> 00:15:42.340 time maybe after the initial filter. Process 00:15:42.340 --> 00:15:45.039 jobs in smaller batches, say 10 or 20 at a time. 00:15:45.179 --> 00:15:47.440 This makes the workflow much more stable and 00:15:47.440 --> 00:15:49.700 manageable, preventing those frustrating timeouts. 00:15:49.879 --> 00:15:52.779 Yeah, and a critical step -by -step testing strategy, 00:15:53.000 --> 00:15:55.570 start small, really small. with just one job 00:15:55.570 --> 00:15:58.110 add a limit node right after your job scraper 00:15:58.110 --> 00:16:00.769 and set it to one pin expensive data like the 00:16:00.769 --> 00:16:03.129 apify scrape results so you don't rerun it constantly 00:16:03.129 --> 00:16:05.809 and rack up unnecessary costs while testing most 00:16:05.809 --> 00:16:08.509 importantly test each node individually methodically 00:16:08.509 --> 00:16:10.870 checking its output did it get the resume did 00:16:10.870 --> 00:16:13.330 it filter correctly did it find an email this 00:16:13.330 --> 00:16:15.590 allows you to isolate problems to a specific 00:16:15.590 --> 00:16:17.929 step rather than you know searching for a needle 00:16:17.929 --> 00:16:21.409 in a haystack crucially always always check your 00:16:21.409 --> 00:16:23.629 gmail drafts before you ever let the system send 00:16:23.629 --> 00:16:25.990 anything automatically manually review every 00:16:25.990 --> 00:16:28.450 single aspect. The recipient's name, the company 00:16:28.450 --> 00:16:30.669 name, the job title, the resume link, make sure 00:16:30.669 --> 00:16:33.669 it works. Give it that final personal human touch. 00:16:34.029 --> 00:16:37.190 And consistently monitor your API usage dashboards 00:16:37.190 --> 00:16:40.490 for OpenAI, Apify, and any mail finder to avoid 00:16:40.490 --> 00:16:43.809 any unexpected cost surprises. Then scale gradually. 00:16:43.850 --> 00:16:46.029 Start with one job, test it, then five, test 00:16:46.029 --> 00:16:48.980 again, then 10, then 25. resolving any issues 00:16:48.980 --> 00:16:51.120 as you go. Finally, let's touch quickly on best 00:16:51.120 --> 00:16:54.299 practices. This system isn't about, you know, 00:16:54.379 --> 00:16:56.779 indiscriminately spamming 10 ,000 jobs you're 00:16:56.779 --> 00:16:59.259 completely unqualified for. That helps no one. 00:16:59.460 --> 00:17:02.620 The AI filter is your ethical gatekeeper here. 00:17:02.759 --> 00:17:05.900 Keep it finely tuned to genuinely find good fits, 00:17:06.000 --> 00:17:09.160 not just any job. This reflects well on you as 00:17:09.160 --> 00:17:12.339 a candidate. And the human review of drafts is 00:17:12.339 --> 00:17:15.019 absolutely crucial for that final personal touch. 00:17:15.400 --> 00:17:18.359 Perhaps you can add one small, nuanced sentence 00:17:18.359 --> 00:17:20.519 the AI might have missed, something specific 00:17:20.519 --> 00:17:23.019 to that company's recent news. And if you do 00:17:23.019 --> 00:17:26.039 choose to use that I built this system line in 00:17:26.039 --> 00:17:28.680 your email, be ready and excited to discuss it 00:17:28.680 --> 00:17:31.119 enthusiastically in an interview. This is your 00:17:31.119 --> 00:17:33.740 genuine chance to shine, demonstrating your initiative 00:17:33.740 --> 00:17:36.559 and technical prowess. Totally. Imagine scaling 00:17:36.559 --> 00:17:38.900 this to, well, okay, maybe not a billion queries, 00:17:38.980 --> 00:17:41.680 but the system, if built right, proves your ability 00:17:41.680 --> 00:17:44.160 to think systematically and leverage modern technology 00:17:44.160 --> 00:17:46.779 effectively. These are precisely the skills modern 00:17:46.779 --> 00:17:48.880 companies are desperate to hire right now. You're 00:17:48.880 --> 00:17:50.400 not just applying for a job, you're demonstrating 00:17:50.400 --> 00:17:52.420 immense value from the very first interaction. 00:17:52.859 --> 00:17:55.119 So what's the most significant advantage of adopting 00:17:55.119 --> 00:17:57.960 that step -by -step testing methodology? It allows 00:17:57.960 --> 00:18:01.200 for precise problem isolation. It lets you find 00:18:01.200 --> 00:18:03.839 and fix issues at each specific stage of the 00:18:03.839 --> 00:18:07.430 workflow, saving tons of frustration. So today 00:18:07.430 --> 00:18:09.829 we've taken a really deep dive into constructing 00:18:09.829 --> 00:18:12.710 an AI system that fundamentally changes the dynamic 00:18:12.710 --> 00:18:15.210 of job hunting. You move from being essentially 00:18:15.210 --> 00:18:18.210 a passive applicant, sending resumes into a digital 00:18:18.210 --> 00:18:21.569 void, to becoming a truly proactive problem solving 00:18:21.569 --> 00:18:23.549 candidate. Yeah, it's all about demonstrating 00:18:23.549 --> 00:18:25.509 immense value from the very first interaction. 00:18:25.950 --> 00:18:28.029 You're not just looking for a job anymore. You 00:18:28.029 --> 00:18:31.450 are creating an undeniable, tangible reason for 00:18:31.450 --> 00:18:33.809 someone to hire you. You're showing them what 00:18:33.809 --> 00:18:36.799 you can do. This AI system is more than just 00:18:36.799 --> 00:18:39.700 a tool to apply for jobs faster. It's tangible 00:18:39.700 --> 00:18:42.599 proof of your ability to innovate and solve complex 00:18:42.599 --> 00:18:45.539 problems using modern technology. It's a statement. 00:18:45.740 --> 00:18:47.660 You now have the complete blueprint, pretty much, 00:18:47.720 --> 00:18:50.480 to forge your job -hunting superpower. The only 00:18:50.480 --> 00:18:53.119 thing left is to actually build it. Take that 00:18:53.119 --> 00:18:56.720 first step. Start small. Iterate, improve, and 00:18:56.720 --> 00:18:59.519 then scale up. And maybe think beyond job hunting. 00:19:00.599 --> 00:19:02.900 What other areas of your life or even your current 00:19:02.900 --> 00:19:05.819 work could truly benefit from a show -don't -tell 00:19:05.819 --> 00:19:08.759 AI system like this? How might demonstrating 00:19:08.759 --> 00:19:11.180 skills proactively rather than just stating them 00:19:11.180 --> 00:19:13.359 fundamentally change other industries, perhaps 00:19:13.359 --> 00:19:16.500 even education or creative fields? Think about 00:19:16.500 --> 00:19:18.259 it. Your dream job might just be an automation 00:19:18.259 --> 00:19:21.119 workflow away. That's all for this deep dive. 00:19:21.299 --> 00:19:23.299 We'll be back soon with another stack of sources 00:19:23.299 --> 00:19:23.819 to explore.