InData And BeyondbyNishant Ambastha·May 24Why Data Engineering Became More Important in the AI EraSomewhere between the foundation model hype of 2022 and the enterprise AI reckoning of 2025, a quiet inversion happened. The teams…A response icon2A response icon2
Nishant Ambastha·May 14Most AI Projects Fail Before AI Even StartsThe real bottleneck isn’t the model. It’s the data operating model.
InData And BeyondbyNishant Ambastha·Mar 18Setting Up Neo4j 5.26 LTS on UbuntuI set up Neo4j on a home Ubuntu machine for local graph database experiments. The requirements were modest — install the LTS version…
InData And BeyondbyNishant Ambastha·Feb 14Netflix Analytics Platform on Azure (Real-Time + Batch Lakehouse Architecture with Databricks &…This project implements a production-grade analytics platform for an OTT platforms (Netflix in this case) — engineered to process millions…
Nishant Ambastha·Jan 28Spark Best Practices(Week 4): Mastering Spark Window FunctionsA Practical Guide to Correctness, Patterns, and Performance
Nishant Ambastha·Jan 19Spark Best Practices(Week 3): Join Explosion, Fact vs Dimension, and Why Schema Design Matters More…This week, I want to connect three concepts that are usually taught separately, but fail together in production:
Nishant Ambastha·Jan 12Spark Best Practices(Week 2): Spark Optimization: Every groupBy is a Shuffle and How to Reduce ItThe Reality CheckA response icon2A response icon2
Nishant Ambastha·Jan 4Spark Best Practices(Week 1): Predicate pushdown + null semantics + deterministic dedupe…Most Spark bugs don’t throw exceptions. They don’t fail your CI/CD pipeline. They don’t trigger alerts.
Nishant Ambastha·Dec 8, 2025Designing Delta Tables That Stay Fast, Trustworthy, and Clean Over TimeDelta Lake is often described as “Parquet with a transaction log”, but that description is too small. Once you work with it in real…
Nishant Ambastha·Nov 20, 2025Azure App Service, Container Apps, and Logic Apps: Think in Roles, Not ProductsWhy This Topic Matters