Document parsing
PDF, DOCX and plain text reduced to a clean text layer, with the file's real bytes checked against its extension before anything reads it.
Parse a résumé, extract its skills against a curated taxonomy, and score it against every open role — in one pass, with the reasoning shown rather than asserted.
Four agents, each with one job and a defined contract between them.
PDF, DOCX and plain text reduced to a clean text layer, with the file's real bytes checked against its extension before anything reads it.
Skills resolved against one canonical vocabulary rather than matched literally, so React, ReactJS and React.js all land on the same skill.
Five weighted rule-based components — skills, title, experience, education, keywords — blended with a trained classifier, each reported separately.
A written rationale per match, tagged with the provider that produced it, falling back to a rule-based writer when no model is reachable.
Live analysis across 0 roles in 0 countries, from 0 companies in 0 cities. Counts from the loaded corpus, not projections.
No corpus loaded — start the API to populate this.
Corpus loaded
0 of 0 markets plotted
0
Countries
0
Cities
0
Companies
Roles by market
Drop in a résumé and watch it scored against all 800 roles — with every component of every score, and the gaps, shown.
Enter the dashboardNothing is uploaded until you press Analyse. Browse the corpus first.