diff --git a/README.md b/README.md
index 9200a9c..5c4d01f 100644
--- a/README.md
+++ b/README.md
@@ -30,7 +30,8 @@ Open in any spreadsheet app. Delete rows as you apply or dismiss.
 - **Paginated scraping** — fetches up to 6 pages per search (150 jobs) with a 400 ms delay between pages to avoid rate limiting
 - **Card-boundary-safe parsing** — each job card is isolated before extracting fields, preventing content from adjacent cards bleeding into company or date values
 - **Language filter** — drops German (`m/w/d`, umlauts, gender-marked vocabulary) and French (`H/F`, accented characters) job titles automatically
-- **Software-role filter** — requires the title to contain a relevant term (software, developer, data, AI, backend, devops, platform, architect, etc.) — eliminates aviation, mechanical, and pharma-process roles that match "Engineer" broadly
+- **Software-role filter** — requires the title to contain a relevant term (software, developer, data, AI, backend, devops, platform, architect, technical, solutions, consultant, product/IT/engineering manager, etc.) — eliminates aviation, mechanical, and pharma-process roles that match "Engineer" broadly
+- **Seniority filter** — drops junior, intern, trainee, and apprentice titles
 - **Global deduplication** — job IDs are indexed from the existing CSV on startup; jobs you've already seen are never re-added, even if you delete and re-run
 - **30-day recency** — only fetches jobs posted in the last month (`f_TPR=r2592000`)