Train-serve skew shown with one feature computed two ways, the point-in-time join that prevents it, and an honest account of when a feature store is overkill.
A feature store exists because the same feature gets computed twice — once in a batch job for training and once in a service for scoring — and the two definitions drift apart. This page shows one feature breaking, in eight lines, and the join that stops it.
One feature, two definitions
The feature is avg_order_value_30d. It appears in a fraud model. The training set is built in the warehouse:
-- training pipeline, run nightly over history
o.order_id,
I rebuilt a tiny feature store from scratch in pure Python to finally understand the point-in-time join: the one operation that…
Part 3 of 5 in the series: When Your AI Pipeline Grows Up In the previous post, we worked through...
ML Feature Store Versioning: Reproducible Quants By Shakti Tiwari · Educational only · Not...
The problem with "just add more workers" Most Spark performance issues on Databricks...
When the autonomous customer‑support bot at Acme Corp crashed after 2 hours, the logs showed a 92 %...
Learn how stitched-together tooling can be a bottleneck for shipping at scale, and how a unified platform can bring together…
