AI agent development studio · Montevideo, Uruguay

AI agents & agentic workflows — built for production. The full spectrum. In one pass.

Two senior engineers. 25 years of software.
We build and integrate AI agents, LLM pipelines and agentic workflows into products that already ship — and leave them lighter than we found them.

Years
25
To prod
8w
Seniors
2

AI Consulting

380–450 nm

Agent Development

450–495 nm

Agentic Workflows

495–570 nm

LLM Integration

570–590 nm

RAG & Knowledge

590–620 nm

Multi-Agent Systems

620–750 nm

00 What we build

Custom AI Agent Development

Autonomous agents with tool use, memory, and multi-step reasoning. Built to run in production — not to demo.

Agentic Workflow Architecture

End-to-end agentic pipelines: orchestration, state management, human-in-the-loop. From prototype to reliable system.

LLM Integration & RAG

Wire language models into your existing stack. Retrieval-augmented generation, fine-tuning, prompt engineering — no vendor lock-in.

AI Consulting & Prototyping

Technical due diligence, architecture review, proof-of-concept in weeks. We tell you what to build — and what not to.

01 Selected work

  • Galileo / SoFi 2024 – 2025
    Fintech
  • Automattic Ongoing
    AI · Open Source
  • La Diaria 2023
    Media
  • Endor 2023
    Dev Tools
  • luz.uy 2025 →
    AI Platform

02 How we work

Ephemeral.

We enter, build, transfer. If you still need us in year two, something went wrong in year one.

Seniors only.

The two people in the kickoff write the code. No pyramid. No handoffs. No surprises.

Speed ships.

A week of ours is a quarter elsewhere. No coordination overhead — just the two of us and your problem.

Full stack.

From AI agent architecture to deploy. One team, one context, one accountable line from idea to production.

Start
a project.

hola@luz.uy
One email. One reply. One call.
Montevideo · Working globally
24h response.