Mindvalley Stories
Case Study · AI Mastery

After an 11-year role ended, they built a five-check filter for their job search

A five-check job filter built in Claude screened 23 real postings, narrowing them to 6 applications and 2 recruiter screens.

After being laid off from a role they had held for 11 years, this job seeker refused to spend their hours on applications that would never fit. They built one simple, consistent filter with Claude and let it surface the facts while they kept the final say. It is a quiet, practical story about turning a hard moment into a clear system.

Monika Jaszi
A professional back on the job market after 11 years in one role, building their own AI tools.

Career Assistant Skill

Before

Laid off from an 11-year role, they faced a job search with no wish to burn hours on postings that were never a real fit. Every promising listing meant researching salary, culture and red flags by hand. Each borderline role invited more hovering and second-guessing.

“After being laid off from an 11-year role, I needed to job search efficiently, without burning hours applying to postings that were never going to be a real fit.”

In their own words

The turning point

They started small. Using Claude's Skills feature, they built one scorecard that runs the moment they share a posting: minimum pay, location fit, red flags in the wording, job level, and a live check on the company's pay, retention and culture. Always five checks, always in the same order. No complicated setup and no new tool to learn.

“Consistency was the real difference-maker, not complexity.”

In their own words

After

They still skim a posting first, then run it through the filter when it looks worth a closer look. So far they have screened 23 real postings and applied to only 6, and two of those became recruiter screens. They apply faster when a role clears and review more postings overall, and they are exploring a template others could adapt.

What it lets people do

The filter gives a job seeker the full picture of a role, not just the posting, in a few moments. It flags vague compensation and unclear scope, checks the company itself, and leaves the decision firmly with the person.

What they learned

They found that consistency, not complexity, made the difference. Keeping the build simple also let them spot problems quickly, like results that did not add up and files that were not real, which a bigger build might have hidden.

“Starting simple is what let me actually notice something was wrong.”

In their own words

Career Assistant Skill was built during AI Mastery.

Read the full submission
Problem Statement:  After being laid off from an 11-year role, I needed to job search efficiently, without burning hours applying to postings that were never going to be a real fit. Achievements:  Built one simple filter, five checks, run automatically the moment I share a job posting: minimum pay requirement, location fit, red flags in the posting itself like vague compensation wording or unclear scope, whether the job level fits my criteria, and a live check on the company itself covering pay competitiveness, employee and client retention, and culture. Since building it, I have screened 23 real postings and applied to only 6 (26%). Two of those 6 applications turned into recruiter screens. Tools: Claude (Skills feature), Claude Design (visual summary poster). Methodology:  The whole idea started small. I still skim a posting first, same as before. But instead of spending time properly researching and weighing a role myself, salary, culture, red flags, all of it, I run it through one scorecard the moment it looks worth a closer look. Five checks, always in the same order: does it meet my minimum pay requirement, does it fit my location requirements, are there red flags in the posting itself, does the job level fit my criteria, and what does a live check on the company show for pay, retention, and culture. That is really the whole system. No complicated setup, no new tool to learn, just one consistent check run every time instead of me manually researching each role myself. Consistency was the real difference-maker, not complexity. The filter surfaced the facts, salary, red flags, culture, but I stayed the one deciding whether to apply. Who it's for and future plans:  Built for my own active job search, and I plan to keep using it as my search continues. I am also exploring turning it into a simple template others could adapt to their own search, since the five-question filter is not specific to my situation. Cost: $0 beyond my standard Claude subscription. Learnings: - I learned that having the full picture, not just the job posting itself, is what let me decide quickly and correctly. - Once a posting cleared, I applied faster and spent less time hovering over borderline calls. That let me review more postings overall, not just decide faster on each one. - Starting simple is what let me actually notice something was wrong. Using it regularly made the problems obvious, results that did not add up, and files being referenced that were not even real. A more complicated build might have buried both of those issues longer. Relevant URLs: N/A, personal tool, not publicly hosted. Attachments: - Full skill zip (SKILL.md, all reference files, the checking tool, README, and the 23-entry tracker) - Career Assistant Skill poster (Claude Design)

Written 7 August 2026 Submitted by the storyteller

Every story is written by a Mindvalley student or by someone our teachers have worked with. How we collect and check stories

More from AI Mastery

Career Assistant Skill: a AI Mastery case study | Mindvalley Stories