Miatz (MEE-ahts) is a founder-led engineering institute. We built the whole engine ourselves — then opened it.
Miatz started with a frustration the founder kept meeting in industry: hiring pipelines full of certificates and portfolios that proved nothing, while the engineers who actually carried systems — who could take a vague problem and own it to production — were vanishingly rare and getting rarer as AI made it easy to look competent without being competent.
So we built the thing we wished existed: a selective, mastery-gated training engine that runs like an athletic program — daily recall, daily reps, weekly retros — wrapped around real work, with an AI shadow engineer (Xook) that mentors and observes, and a record at the end an employer can actually verify. Every feature on this site exists in the product today; nothing here is a mockup.
The institute is run as a founder-led proprietorship, taught with industry experts who take specialist sessions, and operated on the same engineering that builds the platform itself. Small by design. Selective by design. Free for founding cohorts — because footfall and proof come before pricing.
Judgment over output AI makes output cheap. We train and measure the thing that stayed expensive: knowing what to build, what to trust, and what to own.
Everything real Real curriculum, real incidents, real audit logs, real credential cryptography. If we claim it on this site, it runs in production.
Humans set the bar The AI scales the reps; mentors and industry experts decide what excellent looks like.
Learners own their record Your decisions, your data, your credential. You choose what to publish and who verifies it.
Learners: sit Chuun. Companies and experts: hello@miatz.com reaches the founder directly.