About

What this is

Yomu follows the engineering internet for you. It reads the public feeds and sitemaps of teams that write about how they build software, and adds three things to every post: a short summary, the takeaways worth remembering, and topic tags — all generated from the original by a language model. They are concise editorial representations, not independent fact checks, and should be checked against the original before you rely on them.

Under the summaries, everything is searchable, and the Ask endpoint answers questions with citations pointing back at the exact sections it drew from.

How it works

• Posts arrive from each blog’s own public feed or sitemap — nothing is scraped from behind logins.
• The crawler introduces itself by name, honours robots.txt, and paces its requests to stay polite.
• Text is split into sections and embedded, which is what powers search, answers, and the related reading on every post.
• A model writes each summary once; the original is always one click away and always authoritative.

Principles

• Sourced posts are shown as excerpts and a link — the full article stays with its author.
• Source owners are in control: their posts can be changed or removed at any time.
• No accounts, no trackers, no ads. The privacy page lists everything that is stored.
• Every model call is metered; the costs page shows what running this actually costs.

Who

A solo project, built and maintained by Vincent, an engineer who reads more of the engineering internet than is strictly necessary. The fine print lives in the terms and privacy pages.

Contact

For questions, feedback, or other inquiries, email hi@yomu.fyi.