KAS scraper - automated public listing monitoring

Business problem

Manual checks were slow and expensive, while changing website behavior increased the risk of missing critical records.

Approach and solution

I built a modular daemon service that crawls sources on schedule, detects changes, and sends alerts only when new business-relevant data appears.

Delivery scope

• Time-window scheduling with multi-frequency polling strategy.
• Deduplication and idempotent notification pipeline with persistent state.
• Playwright fallback for scenarios where pure HTTP scraping was not reliable.
• systemd deployment with auto-restart, logging, and diagnostic runtime modes.

Business impact

• Relevant updates reach stakeholders close to real time.
• No duplicate alerts and reduced manual operational effort.
• Stable monitoring foundation ready for additional data sources.