Software Engineering Skills for Data Engineers
summary
Covers concrete fixes: pick numeric types to avoid floating point surprises and cut storage (example dataset stored as 22 GB vs ~11 GB when switching doubles to floats), store hashes as binary (32 bytes) instead of hex strings to save space, use pydantic for typed configs, use Databricks notebook code debugger, handle Spark driver vs worker execution and Python GIL pitfalls, resolve a real soda vs dbt core dependency conflict with Poetry, and apply partial application and memoization for reusable, testable pipeline code.
AI-written summary · about
chapters (7)
- 0:00Software engineering intro
- 1:49Data types matter
- 8:31Avoiding spaghetti code
- 17:04Distributed computing
- 22:21Debugging pipelines
- 25:13Managing dependencies
- 27:39Engineering curiosity
related talks
51mPanel Debate: Rise Of The Data Engineer
22mDataEngBytes 2023 - BRI-T1-06 - Yanir Seroussi
40mData Engineering in a Rapidly Changing World
42mBringing Software Engineering Rigour to Data, DataEngBytes 2022
39mWhat Data Engineering Can Learn from Front-end Engineering
57mBACK TO THE FUTURE: EMERGING TRENDS IN DATA ENGINEERING
25m"Data Engineering: The World Between Worlds"
31mFuture of data engineering: what's in store?
