Learn. Question. Feel. — A Product Data Scientist's Field Notes

Writing down what I pick up, what confuses me, and...

How Martingales Make Always-Valid Inference Work

The Fair Game That Powers Always-Valid Inference

Always-Valid Inference

A Gentle Introduction

Bayesian Trees for Sparse Context Prediction

When You Don't Have Enough Data

Statsig Feature Flags

How They Work and Why They're More Than Just an...

Datadog Feature Flags

What It Is and Why Observability Changes Everything

Why Datadog Acquired Eppo

And What It Signals for the Experimentation Market

Feature Store vs. Feature Flag

One for Storage and One for Switch

Feature Store vs. Embedding

Two Concepts That Are Easy to Confuse

TWFE with Staggered Adoption

When Standard Fixed Effects Aren't Enough