Saurabh Choudhary·Sep 26Building a Job Scheduler with Retries, Cancellation, Priorities, and Worker Recovery in Java —…Running a task is easy. Keeping its state correct when concurrency and failure collide is the real challenge.A response icon1A response icon1
Saurabh Choudhary·Sep 11AtomicLong vs LongAdder for Java Request Counters — and When You Need a Rolling WindowChoosing a request counter that stays meaningful under concurrent load.
Saurabh Choudhary·Aug 15Can Public Data Predict a Company’s Next Layoff? I Built a Pipeline to Find OutAnd the results are surprisingly strong.
Saurabh Choudhary·Jul 12Everyone Says “Buy the Dip.” I Wanted to Know How Often That Actually Happens.I tested multiple definitions of the “dip” using 10 years of S&P 500 data — and the results completely changed the way I think about buying…A response icon1A response icon1
Saurabh Choudhary·Apr 15The Memory Wall: Why Bigger Context Windows May Not Improve LLMs’ RecallInformation without reliable recall isn’t intelligence. It’s noise.
Saurabh Choudhary·Apr 4AI Feels Infinite—Until You Look Under the HoodThe biggest constraint on AI isn’t in the model—it’s in the systems we rarely think about.
Saurabh Choudhary·Feb 16HLD: Semantic Cache System for LLM ApplicationsA system design deep dive into semantic cache architecture
Saurabh Choudhary·Jan 18Inside Semantic Caching — Core Concepts: How Meaning Becomes a Cache HitThe architecture, core concepts, and ML decisions behind production-grade Semantic Caching Systems
Saurabh Choudhary·Jan 11Stop Burning LLM Tokens on Repeat Queries — Cache Smarter. Think SemanticA foundational introduction to Semantic Caching — why it’s a necessity, and how it’s both a challenge and an opportunity.