JobShifu proposes changes to your resume, shows the experience behind each one, and lets you review them before exporting. Source checks verify citations. Additional checks flag wording and claims that need your attention.
Named the way recruiters expect, with a short code so you can always tell versions apart.
Tailor, as the app shows it. Every proposed change names the experience it came from. A proposal with no source in your Vault is dropped before you see it, and the finished resume is checked line by line before the file exists.
The Calibration Engine checks resume changes against your confirmed experience. Your Vault is your reusable master resume. When JobShifu tailors your resume for a job, proposed lines cite saved experience. A source check blocks missing or invalid citations. Other checks flag possible issues for review; a valid citation alone does not prove a claim is accurate.
Sound familiar?
Three things an AI resume builder does when it runs out of facts, and what happens to each one here.
It gave me three years of Kubernetes. I have never touched Kubernetes.
A line with no source is never proposed. Proposals citing anything you have not confirmed are dropped before they reach you, and the build fails on any line that slips through.
It rounded "helped with" up to "led".
Scope and title checks flag the stretch. A line claiming more than its source supports, or a title reading above the one you held, is marked on the card. You decide, with the evidence in front of you.
I read it back and it could have been anyone's.
Built only from your confirmed facts, in your words. A resume that can contain only what you did cannot sound like someone else. The AI-tells check flags machine texture on any line that drifts.
How tailoring works: JobShifu proposes, you approve, the check runs last
1
Your Vault holds confirmed facts
Roles, projects, achievements, skills. Import a resume to start it, but nothing becomes citable until you confirm it. Reading a job description surfaces wins you forgot, and confirming one with a tap makes it citable for every resume after.
The model proposes changes, it does not write your resume
JobShifu reads the posting and returns specific edits: keep this line, rewrite that one, drop this, add one built from these Vault entries. Every proposal names the confirmed units it draws from. Proposals citing anything unconfirmed are dropped before they reach you.
add
Senior full-stack engineer with React, TypeScript, Node.js and Python experience shipping user-facing web at scale.
✓ grounded in your experience show source
3
You approve every change, card by card
Each card shows the original, the proposal, and the evidence behind it. Accept, edit, or reject. There is no "accept all" that skips the reading, on purpose: the review is where you catch the line you would not want to defend.
AcceptEditReject21 of 22 reviewed
4
The check runs before the document exists
Your approved changes are assembled and the full Calibration Engine runs on the result. Provenance and job-posting echo pass, or the build fails. Only on a pass does a finished resume exist to preview, download or attach.
They are not equal, and a tool that told you they were would be worth less. Only provenance and job-posting echo are exact enough to fail a document.
Provenance
Fails the build
Every line must cite at least one confirmed Vault entry. No source, no line. Exact, deterministic code, and one of the two checks that stop a resume being produced.
Job-posting echo
Fails the build
A term from the job posting, such as a tool or a standard, that appears nowhere in your Vault or your uploaded resume. Exact, deterministic code, so the same line gets the same verdict every time.
Role integrity
Warns
A bullet attached to the wrong employer or role.
Jagged metrics
Warns
A number that does not match the one you confirmed.
Scope
Warns
A claim that widens what you actually did. Model-backed, so it fails open.
Entailment
Warns
Whether the source really supports the line. Model-backed, so it fails open.
Title seniority
Warns
A title that reads more senior than the one you held. Never blocks.
Positioning
Warns
Framing that oversells the role's remit. Never blocks.
AI tells
Advisory
Phrasing that reads as machine-written. Flagged for you, never enforced.
Two of these (scope, entailment) are model-backed and fail open: if the check itself cannot reach a verdict, it does not invent one. That is why they warn rather than block, and why this page will not describe them as a guarantee.
What happens when a resume fails the check?
You get told which line failed and why, and no document is produced.
Not a warning. The absence of a resume.
A failed build is not a degraded resume with a banner on it. There is nothing to accidentally download, nothing to forget you were warned about, nothing sitting in your documents folder six weeks later that you attach to an application without remembering the banner you dismissed.
A warning you can click past is a warning that eventually gets clicked past, usually at 11pm the night before a deadline. That is exactly when you least want the safety net to be optional.
Does this mean my resume gets worse?
It means your resume stops borrowing from experience you do not have, which is a different thing. Most of what makes a tailored resume work is selection and language: which of your real achievements to lead with for this job, described in the terms this employer uses. Both are fully available inside the constraint.
What it will not do is give you the distributed systems experience the job wants and your Vault does not contain. That line was never going to survive the interview anyway.
Passes the Notepad test
Paste any bullet into a blank document and you can still explain where it came from, because it came from something you confirmed.
Survives the interview
Every line traces to real work, so there is no bullet on the page you cannot talk through for two minutes.
Why do AI resume builders invent experience?
Not because they are badly built. Because of what they are.
A language model asked to make your resume match a job description is given two instructions that quietly conflict. One is "match this job." The other is "stay true to this person." When both can be satisfied, the model rephrases your real work to speak the job's language, which is genuinely useful. When they cannot, something has to give, and the model has no way to know which instruction is the important one.
So it fills the gap. It reaches for the Kubernetes experience the job wants and you do not have. It rounds "helped with" up to "led." It attaches a number to an achievement that never had one. The output is fluent, confident and completely plausible, because fluent and plausible is exactly what the model was trained to produce. It ran out of facts and kept writing.
An AI does not fabricate because it is dishonest. It fabricates when it is under pressure to produce and has run out of things it actually knows.
That framing points at the fix. If fabrication happens when the model runs out of facts, you solve it by giving the model more facts and removing its permission to write without them. Not by asking it nicely in a prompt. The checks that block are not models at all: provenance and job-posting echo are exact, deterministic code reading the citations and terms on each line. The two model-backed checks, scope and entailment, advise rather than block, because a model making judgement calls will sometimes be wrong.
Frequently asked questions
Will an AI resume builder lie on my resume?
Most will, given the chance. A general-purpose language model asked to make a resume match a job description has no way to tell the difference between rephrasing something you did and inventing something you did not. When it runs short of facts it produces fluent, plausible, unverifiable text, because that is what it was built to do.
JobShifu removes the opportunity instead of asking the model to behave. Every line of a tailored resume must cite a confirmed unit in your Vault, and a line with no source fails the build before a document is ever produced.
What does "no source, no line" actually mean?
Every line on a JobShifu resume carries a pointer back to the specific Vault entries it was built from. Before the finished document is assembled, the provenance check verifies each line cites at least one confirmed source and that every cited source exists. If any line fails, the build fails.
This is not a warning banner you can click past. There is no document on the other side of a failed check.
Does JobShifu write my resume for me?
No, and the distinction matters. JobShifu proposes changes and you approve them one at a time. Each proposal is a card showing what it wants to change, what it would become, and which of your confirmed experiences it draws from. You accept, edit or reject each one.
A resume you never reviewed is a resume you cannot defend in an interview, so the review step is deliberately not skippable.
Can I still exaggerate if I want to?
You can edit any line, and JobShifu will not silently rewrite what you wrote. What it will do is tell you when a line drifts past what your own evidence supports. The scope and entailment checks flag lines claiming more than their cited sources contain, jagged metrics flags numbers that appear in no cited source, and the title check flags a display title reading above your Vault seniority.
These surface as warnings rather than blocks, because the line is yours. The one thing that cannot happen is an invented line arriving without you noticing.
Is the calibration check just another AI reviewing the output?
The checks that block are not AI at all. Provenance is exact, deterministic code: it reads the citations on each line and verifies they point at confirmed Vault units. Job-posting echo is deterministic too: it compares each line's terms with the posting, your Vault and your uploaded resume. Neither can be talked out of a verdict, and both return the same answer every time.
Two of the nine checks do use a model, scope and entailment, and those are advisory precisely because a model can be wrong. The guarantee rests on the deterministic checks, not the model-backed ones.
Why does this matter if the resume gets me the interview anyway?
Because the interview is where an invented resume gets discovered. A line you cannot talk about for two minutes is a line that will cost you the offer, and you will not know which line it was until you are sitting in front of someone asking about it.
Every bullet JobShifu produces traces to something you actually did, so there is nothing on the page you cannot walk through out loud.
Will recruiters know I used AI to write my resume?
Sometimes, and the tell is texture rather than tooling: the same buzzwords every other applicant's tool reached for, a claim rounded up past what the candidate can talk through, a line that could belong to anyone. JobShifu builds only from experience you confirmed, in your own words, so the raw material cannot be generic.
Two checks work on the texture directly. AI tells flags buzzword and weasel-word density on any line, and scope flags a line that claims more than its source supports. Both are advisory: you edit the line or keep it, and what a reader meets is your work in the job's vocabulary.
Build a resume you can defend.
Confirm your real experience once. Tailor it to every job, with every line traceable to something you actually did.
Senior Software Engineer, Maps