Devin Thakker

The Common News has reached 800 cities representing more than 72 million people, with coverage expected to reach 1,000 cities by the end of the week.

Why I built CoFinder, the problem I saw with existing cofounder discovery, and why I think founder matching should focus on signal, intent, and actual compatibility instead of volume.

Why local government information is so difficult to organize well, and what I have learned from turning messy agendas, meetings, committees, and local issues into usable product surfaces at The Common News.

What changed as The Common News expanded to 100 cities serving nearly 42 million people: data quality, city-by-city variation, entity structure, publishing systems, and what scale actually means in local civic information.

How I think about the core data model behind The Common News: municipalities, meetings, agendas, summaries, entities, and the relationships that turn messy civic documents into structured product surfaces.

Why local government content rarely wins through broad keywords alone, and how entity-rich civic pages, geographic specificity, and structured publishing make long-tail search the real opportunity.

How I approach trust at The Common News: source grounding, scoped summarization, human-readable structure, and product decisions designed to make AI-generated civic coverage more reliable.

A breakdown of the SEO system I built for The Common News: canonical summary URLs, structured metadata, entity-driven local relevance, server-rendered civic content, and municipality-level internal linking.

A practical build log of creating a public-facing RAG app over large municipal budget PDFs, with source citations, embedded document viewing, quality tuning, and production deployment.

Why I built The Common News, the problem we saw with local government information delivery, and how we are making civic updates easier to follow.