Skip to main content

🛠️ Exporting & Automation

This guide explains how to export job match results, configure your output behavior, and outlines planned automation features like scheduled runs and recurring searches.

✅ Current Capabilities​

At present, Job Explorer allows you to:

  • View match tables directly in the UI.

  • Download CSVs for individual job runs (via the Results tab).

  • Customize the output directory via the config tab (app_config.json).

  • Access all artifacts in data/run_*/ directories, including:

    • 07_final_scored/*.jsonl — structured LLM results
    • 04_serp_jsonl_input/*.jsonl — raw search data
    • *_results.csv — user-facing tables

These outputs can already be integrated into external workflows (e.g., spreadsheets, dashboards, CRMs) manually.

⚙️ Planned Automation Features​

We're designing a lightweight automation layer to help users:

🔄 1. Schedule Recurring Runs​

Run job searches every few days with pre-saved settings.

  • You’d define:

    • Saved job queries (e.g. "Data Scientist remote + Europe")
    • Output format and location
    • Frequency (e.g. every 3 days)

🔧 This would live in a job_queries.json config file, or a dedicated “Recurring Search” panel.

📦 2. Smart Export Bundles​

Automatically group top matches and email or sync to external systems.

Planned options:

  • Export top 10 results per query to .jsonl, .csv, or .md

  • Push to:

    • Google Sheets
    • Notion
    • Slack
    • Email summary

A CLI prototype may look like:

python cli.py export --run-id 20250710T2000 --top-n 10 --format markdown

🧩 3. UI Integration​

A future UI tab may let users:

  • Browse past searches
  • Activate automation per query
  • Choose export format & destination
  • Set default output paths

💡 Developer Note​

Want to contribute to the automation logic?

Check:

Help us shape a flexible and powerful automation layer — ideal for power users, job coaches, and anyone who wants job search to feel automatic.

🧪 Try This Now​

Want a taste of automation? Try:

  1. Running a search with multiple job queries.
  2. Visiting the Results tab.
  3. Downloading the CSV and building your own spreadsheet tracker.
  4. Re-running the app next week to refresh the leads.

Not quite a “cron job” — but still a powerful cycle.

🚧 This feature set is evolving. Early contributions and feedback are welcome.