◎

First posts dropping soon. The blog covers machine learning in practice, economics data, and what working with real-world legal and humanitarian datasets actually looks like. Subscribe to get notified.

Machine Learning

Why Fraud Models Fail in Production — and What to Do About It

Class imbalance is the first problem you read about. Concept drift is the one that gets you in production. Here's a field-level breakdown of both.

~8 min read

Economics

What Sectoral Balances Actually Tell You About Public Debt

The accounting identity behind macroeconomic debates most economists skip in public commentary.

~6 min read

Data Engineering

Building a Quota-Efficient YouTube Data Pipeline

How I reduced YouTube API v3 quota consumption by 98% without sacrificing coverage — the design decisions behind the pipeline.

~10 min read

Policy

Language Access and Data: What the Census Won't Show You Directly

1.7 million people in NYC speak English "less than very well." Here's how to use Census ACS microdata to see what that means geographically.

~7 min read

Visualization

Presidential Instability in Peru: A Data-Driven History

Six presidents in five years. Peru has one of the most volatile executive branches in the Western Hemisphere. Here's what the data looks like.

~5 min read

Newsletter

New posts on ML, economics, and policy data — no schedule, no fluff.

D3.js Visualizations

Interactive data visualizations are coming to MMV Data — built with D3.js directly in the browser.

Request a topic →