Your resume vs. the ATS: an open-source referee

What it does

Resume Matcher ingests a master resume (PDF or DOCX) and a job description, then uses an LLM to rewrite and score your CV for that specific role. It also generates cover letters, highlights keyword gaps, and exports everything to PDF via a drag-and-drop editor with four built-in templates.

The interesting bit

The project runs fully offline via Ollama if you don’t want your employment history leaving your machine. For the cloud-inclined, it abstracts providers through LiteLLM, so swapping from GPT-4o to Claude to a local Llama is mostly a settings change. The database is TinyDB—literally a JSON file—which keeps deployment simple but also tells you exactly where the project sits on the complexity spectrum.

Key highlights

• Local-first option with Ollama; cloud providers include OpenAI, Anthropic, Gemini, OpenRouter, and DeepSeek
• Match scoring and keyword highlighting against pasted job descriptions
• Drag-and-drop resume builder with single- and two-column templates
• Multi-language UI and generated content (English, Spanish, Chinese, Japanese, Portuguese)
• Docker images published for both amd64 and arm64

Caveats

• The README warns the project is in “active development” and features are shifting continuously
• TinyDB as the datastore means this is not built for concurrent team usage or heavy load
• The tech stack table cuts off mid-sentence (“Headless Chromium via Pla…”), so PDF rendering details are unclear

Verdict

Job hunters tired of manually rejiggering bullet points for every application will get immediate value. Teams or agencies needing multi-user workflows, audit trails, or high-volume processing should look elsewhere—this is a personal productivity tool, not an HR platform.