Gonzalo Munoz
GonzaloMunoz

Real World AI Engineering

I'm Gonzalo Munoz, a Cloud/AI engineer who ships production systems and teaches the craft. I judge AI the way a senior engineer does: secure, provable, observable, cost-bounded, governed. Everything here comes from real enterprise delivery, in English and Spanish.

See the work →Read my story
12 AWS certificationsGolden AWS Golden JacketClaude Certified ArchitectAI Academy Enterprise SME & educator11+ Years shipping software and cloud

Who's behind this

Eleven years shipping software. The last few, shipping AI.

I'm a Cloud/AI Engineer delivering AI and cloud systems for enterprise clients, and an SME and educator running an enterprise AI Academy, where I run cohorts that turn engineers into people who can ship AI safely.

I hold every active AWS certification and the AWS Golden Jacket, the community credential you earn by passing them all. I'm also a Claude Certified Architect. But the jacket earns attention; only the work earns authority. Certs tell you I did the reading. The systems I ship tell you I can build.

My angle is the part most AI content skips: what actually holds up in production. Cost per run, prompt-injection surface, traces you can read at 2 a.m., a human in the loop where it matters. That judgment is the difference between a demo and something a team can run on Monday.

I work and teach in English and Spanish, mostly for engineers across LATAM who don't need another hype thread. They need to see the real thing built, with the trade-offs left in.

The receipts

  • 12 AWS certifications
  • AWS Golden Jacket
  • Claude Certified Architect
  • Cloud/AI Engineer — enterprise delivery
  • Enterprise AI Academy — SME & educator
  • 11+ years shipping software & cloud
See every credential →Available to speak?

Signature framework

The Ship-It Scorecard

Before I call an AI system done, it clears five gates. It's the checklist I run in enterprise delivery and the one I teach. It's what separates a demo from something you can run on Monday.

  1. Secure

    Least-privilege, secrets handled, prompt-injection accounted for.

  2. Provable

    Evals and tests that catch regressions before users do.

  3. Observable

    Traces, tokens, and failures you can actually see in prod.

  4. Cost-bounded

    Known spend per run, with ceilings — no surprise bills.

  5. Governed

    Human-in-the-loop where it matters, with an audit trail.

Talks, labs, and the hub where the work lives out loud.

Where I work in public.

Speaking

Talks drawn from real delivery

AI systems, agents, architecture, and certification sessions I can bring to your team or user group.

See topics

Labs

Build with the repo open

Hands-on examples across the modern AI stack, designed to become articles and reusable demos.

Browse labs

Content Hub

Real-world AI, in public

Bilingual labs and field notes on building AI that ships — the same work, documented as I do it.

Explore the hub

Proof I ship — a product I built

Get Certified

A guided, gamified way to actually pass the exam.

The study app I built for AWS and Claude certification prep, from real exam experience. Every concept is one editable file in a versioned tree, the whole app rebuilds byte-for-byte from source, and there are no tracking cookies. Learn seriously, or in pixel mode when you want the game.

  • Six exams covered: AWS Cloud Practitioner, AWS AI Practitioner, and the four Claude certifications.
  • 170 concepts and 543 practice questions, one file each, checked against a frozen oracle so nothing gets silently lost.
  • Two modes: a focused serious track and a pixel-game mode for streaks and momentum.
  • Privacy-respecting by design: analytics from server logs, no cookies.
Open Get Certified

Live at certs.skillrealm.dev · part of the SkillRealm family

Proof I ship — the sibling product

Level Up

The part of a senior engineering career that coding practice can't teach.

A daily brief that picks one concept from a 178-node curriculum, checks whether it landed, and brings it back on an expanding schedule — so you meet an idea again before it fades. Seven domains from foundations to Principal-level judgment, including specialist tracks for AI Engineering and Cloud & Platform. Every concept cites a real source, and the reading list is primary sources only.

  • One concept a day, rotating domains, with the prerequisites respected — miss a day and nothing is lost.
  • Checks you can't click through: build the architecture, categorise the tradeoff, order the sequence.
  • 116 verified primary sources — the paper, the official doc, the post by the team that built it.
  • Bilingual EN/ES, no login, progress stays in your browser.
Open Level Up

Live at levelup.skillrealm.dev · part of the SkillRealm family

12 AWS certifications, the AWS Golden Jacket, and Claude Certified Architect. The reading is done. The work is what matters. See the evidence →

The website is the hub. Everything else points back here.

  • Depth: labs become articles become open repos
  • Reach: talks and community sessions, EN & ES
  • Bilingual field notes across the modern AI stack
  • A permanent, public trail of work that shipped