Job-to-CV matching

Match vacancies with your resume right in the browser — quickly and privately

JDA compares any job description with your resume, highlights gaps, and forms a structured report without sending text to external servers.

100%

analysis stays local or uses your API keys

10

LLM providers available out of the box

<15s

average time to score a vacancy

JDA overlay · linkedin.com00:14

Job Description

Senior Product Manager · FintechExperience with generative models, leading backend teams and launching B2B integrations is required.

Match report

Matches

Automatically highlighted overlaps between description and resume.

Gaps

Missing requirements with links to the relevant lines inside the vacancy.

Recommendations

Actionable suggestions for tailoring your response.

Features

Controls for every detail

JDA gives you full control over models, prompts and integrations. These highlights show why power users rely on it every day.

Floating work window

Drag the overlay anywhere on the vacancy page, highlight blocks manually, and keep context between sessions.

Auto extraction rules

CSS, DOM chain and Template strategies gather LinkedIn, hh.ru, Indeed and niche portals, mixing schema.org/meta/page data straight into the LLM input.

Meta data overlay

Open a draggable panel with every meta tag, search placeholders, copy them, and append captured values to job inputs, results, Notion, clipboard copies and Markdown exports.

Multi-provider LLM

Ten providers including Ollama, OpenAI, Gemini, OpenRouter, Azure OpenAI, Anthropic, Perplexity, DeepSeek, Meta/xAI and Hugging Face with per-model prompts and fallbacks.

Notion export

Map any field, including the new Page meta data source, to keep structured properties in sync with the analysis, clipboard copy, and Markdown notes.

Local privacy

Resumes, prompts and API keys live in the browser storage. Nothing leaves the device unless you explicitly export it.

Workflow

Three simple steps

Each stage stays transparent — from capturing text to the final, shareable report.

1. Capture the vacancy

Use Smart Highlighter or Auto-grab with undo/redo history to experiment with the page.

2. Adjust the context

Pick the right resume, model and prompt template. Global tokens like {{GLOBAL_SYSTEM_PROMPT}} work everywhere.

3. Read the report

See matches, gaps and recommendations. Copy Markdown, export JSON or push the result to Notion in one click.

Flexible rules

DOM chains with named groups, CSS selectors and template mode help parse even the trickiest layouts.

Prompt engineering toolkit

Global and per-model templates, reusable tokens and Markdown formatting to keep responses consistent.

Team-ready presets

Export curated bundles (providers, prompts, integrations) and share them with teammates instantly.

Templates

Ready-to-use JSON presets

Import starter packs for integrations, rules and prompts. Each JSON mirrors the JDA Settings Export structure.

JSON blocks

LLM

Model pools, fallbacks and provider settings.

3 files

Ollama · Llama 3

Download JSON
{
  "providers": [
    {
      "id": "prov_ollama",
      "name": "Ollama Local",
      "type": "ollama",
      "baseUrl": "http://localhost:11434",
      "apiKey": "",
      "timeoutMs": 120000
    }
  ],
  "models": [
    {
      "id": "model_ollama_default",
      "displayName": "Ollama (default)",
      "modelId": "llama3.1:8b-instruct",
      "providerId": "prov_ollama",
      "active": true,
      "systemPrompt": ""
    }
  ]
}

OpenRouter · GPT-4.1 family

Download JSON

Google Gemini API

Download JSON

Ready to work

Install JDA and remove vacancy scoring routine

Install the release from the Chrome Web Store and start matching vacancies right away — no Developer Mode or unpacked folders required.

Privacy first

Your data stays with you

JDA never uploads resumes or vacancies elsewhere. Everything is processed locally or via the APIs you enable. Integrations such as Notion are manual and transparent in exports.

  • Local storage for resumes, prompts and API keys
  • Separate permissions per provider and integration
  • Integrations can be toggled on/off anytime