KK Kartik KumarDATA ENGINEER
OPEN TO WORK
Blog

Writing from the pipeline

Short writeups on data engineering, building pipelines that hold up, and shipping small tools on the side. Written as I go.

Data engineering · 7 min

Taming data skew in Spark

Why one Spark task runs forever while the rest finish in seconds, and how to fix it: spotting skew in the UI, AQE skew joins, salting, and broadcast.

Read note →
Data engineering · 6 min

Building data pipelines that do not break

The habits that keep Databricks and Spark pipelines reliable at scale: idempotency, schema contracts, and failing loudly.

Read note →
Building · 5 min

Why I build Chrome extensions on the side

How a data engineer ends up shipping browser extensions, and what the side projects teach me about building for real users.

Read note →