SHUBHAM PAREEKML Systems · Data Engineering

Research Log / Engineering Notes

Blog

From the Lab

Write-ups on model behavior, retrieval tradeoffs, infra decisions, and the less glamorous debugging that makes systems usable.

Posts

2

Focused essays instead of recycled tutorials.

Coverage

5

Themes spanning LLM systems, data engineering, and career decisions.

Default Lens

Production over novelty

If it does not survive latency, scale, or ugly inputs, it does not count.

Notebook Index

Latest Entry

RAG vs Fine-Tuning: When to Use What

A decision framework for choosing between retrieval and model adaptation under real product constraints.

What You’ll Find

  • RAG vs fine-tuning decisions under real constraints.
  • System design notes from LLM and ETL projects.
  • Honest reflections on what broke and how it got fixed.

Filter by Signal

Narrow the archive by topic. The cards stay deliberately compact so a recruiter can scan the whole index quickly.

2 entries visible
Resume