Resume keyword scanner
for any job description.

Read for weight, not for nouns. A posting tells you what it cares about through three signals: position, repetition, and framing. Anything in the first two paragraphs outranks anything in the final bullet list. Anything mentioned three times outranks something mentioned once. Anything under “required” outranks anything under “nice to have”, even when the nice-to-have sounds more impressive. Once you have that ranking, the keyword set falls out of it: the specific technologies, tools, certifications and job-title nouns attached to the highest-weighted requirements. Ignore the generic filler that appears in every posting and therefore distinguishes nothing.

Yes, but for a different reason than you have been told. Modern systems do handle stemming and synonyms, so the old advice about matching a phrase letter for letter is mostly obsolete. Keywords still matter because of how recruiters use the system: they run searches. Someone filling a role types the two or three terms they care about into the database and works through whoever comes back. If your resume describes the same work in vocabulary nobody searches for, you are technically in the database and practically invisible.

Enough to cover the top-ranked requirements, and no more — in practice five to ten terms, not thirty. The failure mode is not omission but saturation: a skills section listing every technology in the posting reads as unfiltered, and a recruiter who has seen a thousand resumes spots a copied requirements list instantly. A better test than counting is placement. Does the most important keyword appear in the top third of the page, attached to something you actually did?

Yes, with no account. It runs a single cheap extraction call on the posting you paste, and the text is discarded once the result comes back. Nothing is stored.