How it works

Evidence → base resume → application → typeset PDF.

That whole cycle is the part other tools cannot reproduce. Here is each step of it, with what the app actually shows you and why each step is shaped the way it is.

Feed the Career KB, once

Every resume variant you own. The old ones, the role-specific ones, the one that runs to three pages.

Maestro resolves duplicate entities across them, clusters the bullets and builds one Career Knowledge Base. This is the step that keeps paying: everything downstream composes from it, so its quality is the ceiling on everything else.

Bullets that arrive unchanged from a file you wrote are approved on import. Anything merged across resumes, or written by a model, waits in the review inbox for you. Fix a duplicate once and it is fixed for every future application.

The Career KB import dialog consolidating several resume variants into one knowledge base

Build a base resume per track

One per track you actually target. Not one per job.

“New base resume → From Career KB” proposes what belongs and tells you why it left the rest out, with a drafted summary. The instruction box steers shape, not facts: “lead with pipeline and cloud work, keep it mid-level, drop the teaching.”

Bullets are never rewritten here. Approved KB points compose word for word, which is what keeps a generated resume defensible.

The base resume studio with its health grade and a live typeset preview

Capture the job, then pre-scan it

Paste the posting, or grab it with the extension from the board you're already reading.

Score it against every base. Then the knock-out pre-scan checks what the posting actually states — work authorization, OPT policy, salary, years of experience — against your profile.

You find out about a disqualifier before you spend the evening, not at the screening call.

A captured job with its extracted fields and a knock-out pre-scan of the stated requirements

Work the gaps

A gap is a requirement your resume doesn't evidence yet.

The workflow asks pointed questions to surface things that are true but unwritten. Close one and you've either produced new evidence that's yours permanently, or it stays a gap. That's the difference between tailoring and drifting.

Already know the answer? Quick Tailor does it in one request, same honesty rules, no session to open.

Importing KB evidence into a resume, with the drawer showing what is already present

Render it, read it, track it

Cover letter, screening answers, and a real LaTeX or Typst PDF compiled on your machine.

Read the output before it goes anywhere. Your name is on it. Then the tracker carries the application from Saved to Accepted, with inline status, filters for what you did versus what an agent did, and every render filed on disk.

Every AI edit is a per-hunk diff you can revert, so “what changed” is never a guess.

The application tracker, from saved jobs through to signed offers

The rule underneath all five

A gap is closed by evidence, or it stays a gap.

This is the whole discipline. Ask a general-purpose assistant for a tailored resume and you get fluent text that no longer quite matches what you did, in a layout it reinvents each time, with nothing kept between conversations.

Approved evidence composes verbatim

Nothing is invented at create time. Rewriting a bullet is a separate, explicit step that asks you first.

An unevidenced keyword can only reach your skills list

It can never become an experience bullet. That constraint survives Quick Tailor and the agent lane alike.

Every edit is a diff you can revert

Per hunk, on the record, with a resume version behind it. Nothing the AI does to your document is one-way.

Career tooling should be infrastructure, not a rental.

Clone it, run it, keep everything it produces. Nothing here was built to make leaving hard.

Run it locally Read the source

Apache-2.0 · no account · no subscription · your data stays on your disk