About CareerMapAI
Job postings are written as wish lists. Almost nobody meets all of it, and the people most likely to talk themselves out of applying are the ones who’d do the job well.
Free, and staying that way
CareerMapAI runs as a nonprofit. It is free for students and for anyone going after a first role. There is no paid tier, no ads, and nothing in the product that tries to sell you something — there is no billing code in this codebase at all. When we recommend a course, free options come first, and paid ones are labelled.
How the Readiness Score works
Your score answers one question: of the things this job asks for, how much can you already back up? It is not a grade, and not a prediction of whether you’ll be hired. It measures the distance between one posting and one resume.
Scoring each skill
Every skill gets one of five levels, based on the strongest evidence in your resume:
- 0
- Nothing in your resume mentions it
- 0.25
- It's in your skills list, but nothing else backs it up
- 0.5
- You used it in a course or a class project
- 0.75
- You used it in an internship, a job, or a real personal project that went somewhere
- 1
- You've used it across multiple roles, or shipped something with it
A class project counts. An unpaid role counts. The thing you built over a weekend counts. We look for evidence you already have before telling you to learn anything new.
Weighting
A required skill counts twice as much as a preferred one, because missing something under “Required Qualifications” is what actually screens people out. Soft skills count half — postings really do ask for them, but “strong communicator” isn’t something a six-hour course fixes, so we don’t let it dominate.
| Hard skill | Soft skill | |
|---|---|---|
| Required | 2 | 1 |
| Preferred | 1 | 0.5 |
The formula
For each skill with weight w and your level p:
score = 100 × Σ(w × p) / Σ(w)Add up what you’ve earned, divide by what was available. The score is calculated by code, not by an AI — the model reads and classifies, the arithmetic is fixed and always gives the same answer for the same inputs.
If we got something wrong
You can remove any skill we picked up and re-map any we matched incorrectly. Removed skills leave the calculation entirely — they come out of both the top and the bottom of the fraction, so removing something you can’t do never penalises you. Your corrections stick, including when the analysis is re-run.
What it can’t tell you
- It scores one posting against one resume. A different posting for the same job title can give a different number.
- It reads what your resume says, not what you know. If you’ve used something and didn’t write it down, we can’t see it — and writing it down is often the fastest way to raise your score.
- The weights are a starting point, not a finding. We’ll tune them against real data.
Your data
Your resume file is read and then deleted — we keep the parsed details, never the upload. Resume text is sent to one AI provider to be structured, and to nobody else. It is not used for advertising, not sold, and not shared.
Ready to see where you stand?
Map your first job