I Built an AI-Powered Job Application System (And It Actually Worked) 🎯

If you're job hunting right now, or quietly keeping your options open, you'll know how broken the process feels. Hours spent tailoring CVs, writing cover letters, chasing applications. And then... nothing back.

I've been there. So I built a system: an AI-powered, end-to-end workflow that covers everything from your master CV to tailored applications in minutes, right through to debriefing your interviews in real time.

It changed my results. Here's exactly how it works. 👇

But First: Is AI Making Job Hunting Worse?

With more applicants using AI, more automated screening, and more noise in general, it's a fair question. My answer's at the end of this post. Let's get into the system first.

I'm using Claude for this, but the same principles apply to whichever AI tool you prefer.

Step 1: Set Up Your Foundation 🗂️

Before you apply for anything, get clear on what you actually want:

✅ The roles you're interested in
✅ Your salary range
✅ Permanent, contract, or fixed-term?

Write it all down as a memory file inside your AI project. It sounds obvious, but having this defined means every decision after this point is faster and cleaner.

💡 Grab my Target Role Brief template to get this set up in minutes.

Build your Master CV

This is the important bit. Your master CV isn't the CV you send to employers. It's a comprehensive document containing:

Every role you've ever had
Every responsibility and achievement
Every skill, tool, and area of knowledge from your entire career

It should be longer and more detailed than anything you'd normally send out. Think of it as your source of truth. Once it's done, add it to your project's memory too.

Why do this first? Because everything else flows from it. Every tailored CV you create later gets pulled from this master. You're never starting from scratch again.

Set your cover letter rules

Same logic applies here. Decide once:

How long it should be
What sections must always appear
What tone it should take

Write these rules into a separate markdown file in your project. Claude will apply them every single time, no exceptions.

💡 I've put together the Job Application Template and Cover Letter Rules you can copy straight into your own project, prompt included.

Step 2: Build Your Pipeline Tracker 📊

You need somewhere to manage your applications. I use Notion, but a spreadsheet works just as well.

Set up a database with columns for:

Company name
Job title
Salary range
Days in the office
Application status
Any other comments

Keep status simple: Applied, Screening Call, Interview, Rejected.

Most of these details get populated automatically when Claude creates your CV for a role. The only thing you update manually is the status column.

If you're juggling 10 or 15 applications at once, having this live and up to date means you're never confused about where things stand. ✅

Step 3: The Application Workflow 🤖

Here's where the system really earns its keep.

1. Paste the full job description into Claude. Claude assesses the fit, flags any hard requirements, and highlights significant skill gaps between you and the role.

2. If the gap's too big, skip it. No CV. No cover letter. No time wasted. This alone saves hours, because Claude is essentially coaching you on whether a role is even worth applying for.

3. If it's a good fit, Claude tailors your CV. Pulling from your master CV, it adjusts your profile summary, key skills, and experience bullets to match the language and priorities of that specific job. This matters enormously for the automated screening software most companies use before a human ever sees your application.

⚠️ Important rule: Claude should only surface things that actually exist in your master CV. Don't let it invent qualifications or experience you don't have.

4. Claude writes a fresh cover letter every time. Copy the output into wherever you keep your CV (I use Google Docs, since AI still isn't great at designing an aesthetically pleasing CV layout).

Step 4: Interview Prep and Reflection 🎙️

Once the CV and cover letter are sorted, logging happens automatically. Inside your Claude project, create a brand new chat for every single application.

Why separate chats? The project holds your master CV and core content, but each application and interview gets its own dedicated space. That keeps feedback from different roles from getting mixed together.

Before the interview

Go into that job's chat and ask Claude to prep you:

What questions are they likely to ask?
What are the priorities of this role?
What should you be emphasising from your background?

All done in minutes.

During the interview

I use Granola to record the call in the background, listening in via system audio. That means I'm fully present, not scribbling notes or trying to remember what they said about team structure or success metrics.

After the interview

Ask Granola's AI chat directly:

How well do you think I came across?
How was I received?
What's the likelihood I get a second interview?

It's a genuinely useful outside perspective on a call you were too busy being present for to properly judge yourself.

For the next round

Since Claude is connected to Granola, it can prep you for the next stage using full context from the last interview. Referencing details the interviewer mentioned last time genuinely makes an impression. 👀

Does It Actually Work?

This system has helped me land a job. It's not magic. You're not going to find a job tomorrow because of it. But it's genuinely helped me navigate a tough market, and I got noticeably more positive feedback on applications compared to before, when I was just firing off generic CVs and hearing nothing back.

So, Is AI Making Job Applications Worse?

Yes, if everyone's doing it the same lazy way. If you're just asking ChatGPT for a generic CV and firing it at 50 roles, you're adding to the noise, not cutting through it.

This system is built to do the opposite:

✅ A master CV that keeps everything honest
✅ A fit assessment, so you only apply where you've genuinely got a chance
✅ Tailoring to the precise needs of each role
✅ An interview prep and reflection loop that makes you sharper every time

Use AI properly, and it makes sure the best version of you actually gets seen. That's the difference between a follow-up interview and your application being binned.

The Bigger Pattern

This is really just one example of a loop I use everywhere: capture, process, store, communicate, automate. Same principle whether it's job applications, content creation, or day-to-day admin.

Once you start seeing your tools as a connected stack rather than a pile of separate apps, everything gets faster. 🚀