Skip to content

Built to ship.

We build awesome, useful products. AI for speed, humans for the craft.

15+
years building
2+100s
humans + agents
daily
ship cadence
2025
founded · miami
100%
super amazing
The work

We ship our own.

Designed, built, and run in-house. Most live today, one open-sourced, one folded in.

What we do

We'll ship yours too.

AI-native apps, automation, and agents. Same hands, same speed.

01

Apps

AI-native web and mobile apps built for real users. Designed with care and shipped fast: the ones above, plus more on the bench.

02

Automation

Systems that connect, automate, and scale: the AI pipelines quietly powering everything we ship.

03

Agents

AI agents and open-source tooling that extend what a small team can do. Built for ourselves first, shared when it makes sense.

Tell us what you're building

We take on a few client builds a year, when the fit is right.

The studio

A small human team
with a bench of AI agents.

mx.works is an independent product studio. We design, code, and run our own AI-native apps, plus the occasional client build when the fit is right.

No layers, no handoffs, no meetings about meetings. Small by design, fast because of it. Everything on this page was designed, built, and shipped in-house.

Read the notes →
“I've shipped with fifty people and I've shipped with two of us plus a bench of agents. Same standards. Way more fun.”
Max, founder of mx.works Max founder · mx.works
The mx.works studio: a long wood table with pink chairs and the (mx) mark on screen
Studio IDmx.works
Founded2025
BaseMiami
Humans2
AI agentsseveral
Apps live6
The stack

Frontier models,
proven stack.

Claude, GPT, and Gemini carry the fast work. Rails, Postgres, and n8n hold it up. We pick tools we can still reason about at 2am, so what ships is something we can fix, not just something that ran.

We open source a good part of what we build for ourselves: the agent scaffolding, the internal tooling, the design system this page runs on.

Have a look through the public repos — mx2, the studio tooling, and whatever else was worth sharing. Fork it, break it, take what's useful.

View on GitHub
OpenAI
Anthropic
Google Gemini
Figma
Ruby on Rails
n8n
Notes

Building in public.

What we're building, exploring, and learning. Occasional updates from the studio.

FAQ

Before you ask.

What is mx.works?

An independent studio where we build and run our own AI-native products. Small team, fast iteration, products we'd want to use ourselves.

What kind of products do you build?

AI-native tools, productivity apps, and automation systems. We focus on problems we understand deeply and products that can compound over time.

What makes your apps different?

Simplicity and craft. No feature bloat, no dark patterns, no subscription traps. Just tools that do one thing well and get out of your way.

What tech do you use?

OpenAI, Anthropic, and Gemini models for AI. React and Rails for apps. n8n and custom pipelines for automation. Figma for design. We pick whatever fits the problem, not the other way around.

How fast do you ship?

Fast. Most products go from idea to live in weeks, not months. Small team, zero overhead, and AI in every part of the process. That adds up.

How can I follow along?

Subscribe to our notes for updates on what we're building and learning. We share openly and regularly.

status: shipping · next slot open

Let's explore
your idea.

Bring the idea, rough is fine. The call is with the two humans who'll design and build it, and you'll leave with an honest read on scope, speed, and fit.

Let's chat

same-day reply · no deck required · bring the itch