A noise sieve — scattered sources become actionable reports.
01 — The problem
Relevant public web content was spread across inconsistent page structures and feeds; manual monitoring, extraction, and deduplication did not scale.
02 — The system
Select a layer to see its role.
Data flow
Technology — why it was used
Python
Scrapy
Selenium
AWS Lambda
Amazon S3
EC2
Terraform
Bitbucket Pipelines
05 — Evidence

Takeaway
Fragmented inputs need visible failure on schedule — not silent drift.
What landed