Martin BrianSenior AI Engineer

Let’s build something that ships.

I ship agentic AI that engineering teams actually adopt. Not demos. Not pilots that quietly die. Things that stay shipped.

psst — try the ? key

~/.claude/skills/martin-brian
YAML · 14 linesbooting

I turn fuzzy ideas into tools people actually use. Not demos. Not decks. Things that ship and stay shipped.

My background is a weird mix — engineering, design, innovation labs, teaching, and a bookshop startup. Turns out that combination is exactly what you need to build AI that works for real humans in real orgs, not just in notebooks.

Ship, then polish

The best tool is the one people are already using. I'd rather put something real in front of someone tomorrow than something perfect in front of nobody next quarter.

AI is a means, not a flex

I don't build AI because it's cool. I build it because some tasks are too boring, too slow, or too error-prone for humans to keep doing by hand.

Understand the room

Design background, innovation labs, teaching. I've learned to talk to stakeholders, not just compilers — which is usually where "AI projects" quietly die.

Languages

Spanish · English

Education

Universidad de Montevideo · Coderhouse

Credentials

DeepLearning.AI · Cambridge C2

SKILL.md

Reading the skill card out loud

The terminal up top showed the spec. Here’s what it means in practice — when to invoke me, and what you get back.

invoke
: martin-brian
handles
: fuzzy idea → production
returns
: a thing that ships
Triggers
  • whenA team has an AI idea and no clear path from prototype to production
  • whenSomething agentic needs to survive real infra — latency, reliability, cost
  • whenStakeholders need explaining as much as the code needs writing
Side effects

May cause people to stop asking “does AI actually help?” and start asking “how did we work without this?”

The story so far

Now

Marvik → dLocal

Making AI do the boring parts

Embedded in dLocal's AI team, where I ship agentic systems on a Claude Agent SDK / Claude Code base. The one that rolled out across the engineering org drives 40%+ end-to-end automation of dev tasks on a high-availability fintech platform. People stopped asking "does this help?" and started asking "how did we work without it?"

Claude Agent SDKPythonMetaflowKubernetes
Four years

Ulta Beauty · Innovation

AR, AI, and a Fortune 500

Contributed to GLAMlab — Ulta's AR/AI virtual try-on and its first public generative AI app. Built a video analysis engine that watches affiliate and influencer content for brand-guideline violations at scale. Plus supporting iOS, web, and backend work across innovation projects.

Google ADKComputer VisionSwiftReact
Alongside

BroLi · Co-founder

Running a bookshop with code

Co-founded and led engineering at an independent bookshop that applied blockchain and AI to a category that usually ignores both. Owned the stack end to end, led a small team, learned that startups are 20% code and 80% everything else.

BlockchainNLPTeam Lead
Earlier

Universidad de Montevideo · Octobot

Teaching, shipping, learning

Taught innovation and life-design workshops at UM's corporate innovation lab. Before that, shipped Django/React web and Swift/SwiftUI iOS at Octobot. The teaching part stuck — I still think of great engineering as mostly good explanation.

TeachingiOSFull-stack
I think in
Python·Swift·TypeScript
I build agents with
Claude Agent SDK·Claude Code·LangGraph·Google ADK
I make AI useful with
RAG·Knowledge Graphs·NLP·Computer Vision
I ship on
Kubernetes·Metaflow·Docker·Vercel

Things I've built

The projects that taught me the most — or that I just can't stop talking about.

40% Dev Automation

The system that changed how an entire engineering org works. Model-agnostic agents that handle code reviews, test generation, and routine development — all running in production on fintech rails.

Claude CodePythonKubernetes

GLAMlab

Ulta Beauty's AR/AI virtual try-on. Millions of users on iOS and Android, and the company's first publicly-recognized generative AI application.

AR/VRSwiftiOS

Brand Compliance Engine

Watches thousands of influencer videos and flags the ones that break the rules — mentions of competing brands, quality issues, policy violations. Built to replace a team of humans squinting at screens.

Computer VisionNLPPython

Let’s connect

Open to conversations about AI engineering, agentic systems, and interesting problems worth solving.

hello@martinbrian.com