Writing

Latest Insights

Deep Learning 2026-04-14 12 min read

Attention Is All You Need — Code & Paper Analysis

A mapping between a minimal GPT-style implementation and Vaswani et al. (2017), plus key conceptual explanations for positional encoding and self-attention.

Read Article

Tech 2026-04-10 9 min read

The Impossible Shrink: How Google's Gemma 4 and TurboQuant Just Redefined Open-Source AI

Gemma 4, TurboQuant, and why high-tier AI can finally run local—without data-center GPUs or restrictive licenses.

Read Article

Tech 2025-12-14 5 min read

Go Big, Then Go Small: LinkedIn's AI Recommendation Engine

A dive into new species of Recommendation Engines

Read Article

Tech 2025-12-14 5 min read

Beyond the Binge: Nerflix's Recommendation Engine

Netflix's big bet on unifoed Recommendation model

Read Article

Tech 2025-12-14 5 min read

4 Surprising Lessons Instacart Learned When Using LLMs to Fix Its Search

Search is a complex problem, and it's not always easy to get right.

Read Article