Daily job posting watcher into Notion with a Slack top 3

Every day at 9am, I want an agent that watches public job boards for new roles matching my sourcing brief, files them in Notion, and posts a Slack digest of the top matches. Trigger is a daily cron at 9am in my local timezone.

Step 1. Run ScrapingBee Google Search across a configurable list of role variants and locations. Queries should look like: site:jobs.lever.co OR site:boards.greenhouse.io "Senior Backend Engineer" "New York" and one variant per configured (role, location) pair. Collect all organic result URLs from each query, dedupe within the run.

Step 2. For each unique result URL, call ScrapingBee Scrape HTML to fetch the listing page (render_js as needed), then ScrapingBee Extract Data with AI (ai_query) to pull a structured object: title, company, location, salary range, posted date, apply link, and a short relevance_score from 0 to 100 against my configured ideal-candidate spec, plus a one-sentence rationale. Skip any posting whose relevance_score is below the configured threshold (default 70).

Step 3. Before writing to Notion, dedupe against my existing tracker by querying the Notion database (data source) for any page whose Apply URL property matches the candidate posting. Skip anything already tracked.

Step 4. For each genuinely new posting that cleared the threshold, Create a Page in my Notion job-postings database with properties for title, company, location, salary range, posted date, apply link, relevance score, rationale, and the date discovered.

Step 5. End with a Slack Bot Send a Message to my configured channel: a digest of the top 3 highest-scoring roles found today, each line showing title, company, location, score, and a link to the Notion page. If nothing cleared the threshold, post a one-liner saying no new matches today.

Make these configurable at the top of the workflow: the role variants and locations to search, which job boards (site: filters) to include, the ideal-candidate spec used for scoring, the relevance threshold, the Notion database id, and the Slack channel.