The problem

A résumé that doesn't speak the language of the job post rarely gets past an applicant tracking system, and rewriting it for every application takes hours. Early-career candidates and people switching fields feel it most: they have the skills, but not the words recruiters search for.

What Darzi does

Darzi ("tailor" in Hindi) fits a résumé to a specific job. It parses what you already have, reads the posting, and rewrites summaries and bullet points around the role's real requirements, without inventing experience.

  1. Your résumé
    PDF, DOCX or text; or import from LinkedIn and GitHub
  2. Parse
    Local extraction first, Gemini for the hard parts
  3. Tailor
    Rewritten against the job post's keywords
  4. Output
    ATS score, suggestions and a LaTeX résumé
  • Hybrid parsing: a local parser handles well-structured files quickly; Gemini fills in the messy ones, so the structured data is reliable either way.
  • ATS optimisation: keyword analysis against the posting, with a score and concrete fixes.
  • Templates and LaTeX: pick a template and get typeset LaTeX output, with a live editor for final touches.
  • Accounts: Clerk sign-in, with a dashboard to start a new résumé or improve an existing one.

How it's built

  • Frontend: Next.js and Tailwind CSS.
  • Parsing API: FastAPI with PyPDF2 and python-docx, calling Gemini where local extraction falls short.
  • Agent tools: a FastMCP server, so the same résumé tools are available to MCP clients.
  • Deployment: Docker images for each service.
Darzi's landing page: Align your resume with the job, instantly
The landing page