Articles

Selected writing on AI, ML systems, and the engineering around them.

2026

  1. Production LLM Inference on Google Cloud

    A production design for serving open-weight LLMs on GKE, covering routing, KV caches, autoscaling, rollouts, failures, and cost.

  2. Speculative Decoding

    How speculative decoding preserves a target model's distribution while replacing serial decode steps with parallel verification.

  3. Model Compression via Knowledge Distillation

    How knowledge distillation trains a smaller model from a teacher, including the core loss, practical variants, code, and common failure modes.

2025

  1. Low-Rank Adaptation (LoRA)

    Why LoRA works, what its factorization changes, and the targeting, optimization, quantization, merging, and serving traps that matter.

  2. Mixture of Experts - Mathematical Foundations and Scaling

    How sparse Mixture-of-Experts layers route tokens, balance expert load, scale capacity, and trade computation for communication.

  3. The Equations That Changed The World

    A fun, increasingly intimidating tour of seventeen equations that reshaped science, engineering, computing, and finance.

  4. AI and the Art of Subtle Control

    How ranking and prediction systems shape attention and choices, where claims outrun evidence, and what meaningful user control looks like.

  5. Data Pipelines And Evolution of ETL vs ELT

    A production-minded guide to contracts, replay, event time, idempotency, observability, backfills, and choosing between ETL and ELT.

  6. Understanding the Internal Mechanics of LLMs

    A grounded tour of tokenization, transformer blocks, causal attention, training, inference, KV caches, scaling, and quantization.

  7. Python Memory Management

    How CPython allocates, retains, and releases memory—and how to measure real memory problems before trying to optimize them.