Blog
Notes from the data floor.
Data engineering, AI, and platform design — written by people who run pipelines in production.
May 9, 2024
Integrating Advanced ELT Pipelines for Data Warehouse Optimization
Why ELT beats ETL at scale, and five practical techniques — incremental loading, parallelism, caching — to keep your warehouse fast and affordable.
Data Engineering · Datablast · 3 min read
Apr 30, 2024
Practical Guide to Implementing a Single Source of Truth
Why a single source of truth changes how organizations decide, a pragmatic path to building one, and the two decisions that actually stall the project.
Data Strategy · Datablast · 5 min read
Apr 19, 2024
Automated Data Movement: What Changes When Nobody Is Watching the Transfer
Automating data movement does not remove failure — it changes its shape. Frequent, loud and cheap becomes rare and silent. What breaks, and what catches it.
Data Engineering · Datablast · 5 min read
Mar 20, 2024
ETL vs ELT Is the Wrong Question
Every comparison table reaches the same verdict. The decision that still matters is where you draw the transform line — and there are three places, not two.
Data Engineering · Datablast · 5 min read
Feb 23, 2024
Data Quality Improvement: Detection Is the Easy Half
Queensland Health went live on a payroll system with 2,400 known defects and paid 78,000 people wrong. The checks were not missing. The authority to stop was.
Data Quality · Datablast · 5 min read
Dec 14, 2023
What Is Data Mesh? The Four Principles, and the Organization They Assume
Data mesh has four principles and one prerequisite nobody quotes. What each asks of you, and the test for whether it was written for your company.
Data Strategy · Datablast · 4 min read
Sep 14, 2023
Game Analytics: The Bottleneck Is the Join, Not the Data
A twenty-title portfolio needs forty analyst-hours to look at once, because attribution, in-game events and ad revenue live in three systems that never meet.
Data Analytics · Datablast · 5 min read
Sep 11, 2023
How to Use Google BigQuery: The Four Decisions the Tutorial Skips
Loading a CSV and running a query takes ten minutes. This is about the four decisions around those ten minutes — two of which you get one chance at.
Data Engineering · Datablast · 5 min read
Aug 24, 2023
Data Integration: The Nth Source Is the Expensive One
Connecting two systems is a morning's work. Connecting twelve is a permanent job — and the thing that actually breaks is never the connection.
Data Engineering · Datablast · 4 min read
Aug 16, 2023
Data Observability: Why Watching Your Data Isn't Enough
The five pillars watch what arrives in your warehouse. None watches the job that was supposed to deliver it — where the quietest failures start.
Data Engineering · Datablast · 5 min read
Aug 4, 2023
Choosing an ELT Tool: The Criteria That Don't Differentiate
Connectors, scale, security, ease of use: none of them is why teams regret the choice eighteen months later. Five questions that actually separate tools.
Data Engineering · Datablast · 5 min read
Jul 27, 2023
Cohort Analysis: Why a Blended Retention Rate Tells You Nothing
Two cohorts from the same year, thirteen points apart by month eight. Average them and you get a number that describes neither, and moves on cohort mix alone.
Data Analytics · Datablast · 5 min read
Jul 19, 2023
7 Benefits of a Cloud Data Warehouse, and the Condition Attached to Each
Every benefit of a cloud data warehouse is real. None is automatic. What each one actually depends on, and why the first bill surprises so many teams.
Data Engineering · Datablast · 6 min read