AI systems builder and product operator

Liang Yuxin

I build AI systems and product workflows that make complex work clear, repeatable, and genuinely useful.

  • Product judgment
  • AI systems
  • Operating clarity

Evidence

Selected work

Only work that raises the average stays.

See all work
  • Loading selected work

Approach

Three principles. One standard.

  1. 01

    Judgment before output

    Find the constraint that matters before adding scope, features, or motion.

    Problem framing / Product strategy

  2. 02

    Systems over features

    Connect product judgment, engineering, and the operating loop people actually follow.

    AI workflows / TypeScript / Cloudflare

  3. 03

    Evidence over claims

    Make outcomes visible and preserve the decisions that turn one result into reusable knowledge.

    Evaluation / Data modeling / Operations

Source of truth

See the work behind the claims.

Explore GitHub