Quick Start #
Five minutes from zero to your first tailored application drafts.
1. Setup #
cd raven
./raven setup
2. Configure filters and profile #
Edit config/portals.yml — what jobs to find:
title_filter:
positive:
- "software engineer"
- "backend"
- "developer"
negative:
- "nurse"
- "sales"
location_filter:
allow:
- "Remote"
- "India"
Edit config/profile.yml — who you are when drafting:
identity:
name: "Your Name"
email: "you@example.com"
location: "City, Country"
links:
resume: "files/resume.md"
portfolio: "https://yoursite.dev"
github: "https://github.com/you"
linkedin: "https://linkedin.com/in/you"
Replace files/resume.md with your real resume (Markdown with ## Experience and - bullets works best).
3. Discover jobs #
./raven discover --since 7
Raven loads filters from config/portals.yml automatically. Results save to data/jobs.json.
For a faster first run (board feeds only):
./raven discover --sources boards --since 7
4. Draft applications #
./raven draft --max 25
Outputs:
drafts/outreach-YYYY-MM-DD.csv— spreadsheet for review and senddrafts/outreach-YYYY-MM-DD.md— human-readable review
5. Apply #
application_type | What to do |
|---|---|
| form (Ashby, Greenhouse, Workday, …) | Follow form_steps in the CSV or .md file |
Fill contact_email, then raven send --dry-run |
Always review drafts yourself before submitting or sending.
Next steps #
- Example dry run — detailed walkthrough
- Discover command — all flags
- Profile configuration — resume parsing and Gemini