DATA · CLOUD · AI · THINGS LEARNED THE HARD WAY
I’m Pravin. I write about the practical side of modern data platforms — architecture decisions, migrations, Databricks, Azure, AI experiments, cost, security, and the lessons that rarely make it into documentation.
NOTES
Short, useful write-ups from real engineering problems — without turning every lesson into a 20-minute read.
The controls that matter when moving notebooks and jobs at scale: dry runs, validation, rollback, auditability and safe overwrite protection.
First article planned →A practical framework for comparing cost, concurrency, operations, governance and migration effort instead of benchmark numbers alone.
In the backlog →Idle compute, warehouse sizing, overlapping workloads and the telemetry worth looking at before asking teams to “optimize cost.”
In the backlog →BUILDING IN PUBLIC
Small tools and experiments that turn repetitive engineering work into something reusable.
Utilities, experiments and reusable building blocks for Databricks platform work.
View on GitHub ↗Experiments and tooling around Synapse-to-Databricks migration patterns.
View on GitHub ↗A useful problem will earn this slot. No fake project cards just to fill the grid.
In progressABOUT
Bits of Pravin is where I document ideas that survive contact with real systems.
I work around data platforms and cloud engineering, with a strong interest in how architecture, security, operations and cost collide in production.
This site is intentionally focused on useful observations, repeatable patterns and tools. Some posts will be polished. Some will simply capture a lesson before I forget it.
Follow what I’m building on GitHub ↗BITS OF PRAVIN