/amplitude-taxonomy-errors-etl

Automation of extracting data from Amplitude taxonomy error report emails and loading data into Big Query for ongoing reporting.

Primary LanguagePythonMIT LicenseMIT

License: MIT Python

Extract–Transform–Load (ETL) pipeline that ingests Amplitude Taxonomy Validation Error notification emails from an IMAP inbox, parses the error tables, and loads them into BigQuery for monitoring and reporting.

Why this project?

Amplitude surfaces validation errors (unplanned events, invalid properties, etc.) in its UI and email notifications. Those errors are valuable for:

  • Monitoring data quality trends over time
  • Automating alerts for broken instrumentation
  • Giving product & data teams visibility without manually checking Amplitude’s UI

This ETL automates that process:
Fetch → Amplitude error emails via IMAP
Parse → extract error tables & timestamps
Load → append to a BigQuery table for analysis & dashboards

Run locally:

python -m venv .venv && source .venv/bin/activate
pip install -e .
amp-tax-etl