Data Engineer, Ingestion & Change-Data-Capture

We take in messy customer files from dozens of partners and turn them into one clean, canonical customer table with a full history of every change. We're hiring a data engineer to own that ingestion and merge layer: validating raw records, catching the bad ones, and keeping an accurate history as records change over time.

What you'll do:

• Build a correct slowly-changing-dimension (Type 2) merge, so re-running the same batch never duplicates history and effective dates never overlap or leave gaps
• Coerce raw text into real types without losing meaning, so a leading-zero ZIP stays a string and money stays exact
• Quarantine bad rows with a clear reason instead of dropping them silently
• Keep dedup and output deterministic, with the current time passed in rather than read from the clock

What we're looking for:

• Solid Python and careful data modeling
• Experience with dedup, upserts, and merge logic, plus the null-handling bugs that come with them
• The discipline to keep a transform simple instead of over-building it

Nice to have:

Background in data pipelines or warehousing, comfortable handling money with Decimal, and mindful of idempotency