[ available for freelance ]

I build

Senior Creative Technologist at .Monks — 15+ years making machines creative, interactive, and human.

Generative AI LLM Pipelines Real-time Visuals Creative Coding Machine Learning
View Work ↓
$ ls ai-projects/
lider-omnicom/
gemini-experience/
ai-chef-system/
stable-diffusion-photobooth/
bk-content-pipeline/
$ cat skills.txt
→ LLMs · Stable Diffusion · ComfyUI
→ Python · JavaScript · TouchDesigner
→ Real-time interaction · AR · XR
$ ▊
work shipped for
Google Walmart Chile Burger King Adidas Pharmaton Charité
00 / latest

Lider × Omnicom

Four AI-generated Instagram pieces for Walmart Chile — no cameras, no 3D, no set. Just reference images and the right prompts.

client
Lider (Walmart Chile)
agency
Omnicom Media Chile
format
Vertical 9:16 · Instagram
pipeline
Image refs → AI generation
canela · 01
Canela at the checkout. My actual cat as a Lider character — built from real photos and placed inside the store through prompting alone.
canela · 02
Canela in the aisle. Same character, different context — visual consistency achieved without a single 3D render.
ciby · 01
Ciby in action. Lider's official brand character, reconstructed in AI using past campaign imagery as the visual reference.
ciby · 02
Ciby in store. Same style guide, different scene — keeping a brand character coherent frame to frame.

Creative

Canela started from photos of my own cat — used as reference to build a new character inside the Lider universe. Ciby is the brand's official mascot: I worked with NanoBanana using past campaign imagery to define consistent hero frames before generating anything. No actors, no set, no camera.

Technical

The starting point was always real reference images — photos of the cat, frames from past campaigns. From there: image synthesis to lock the look, then AI video generation to bring them to life. Post production handled subtitles, motion graphics, and audio.

01 / selected work

AI Pipelines & Experiences

Building real experiences where AI does something genuinely useful — or genuinely surprising.

Google Gemini Immersive Experience
interactive installation · real-time AI

Google Gemini Immersive Experience

A physical installation in New York where Gemini responds to visitors in real time — AI output driving generative visuals live, not pre-rendered. Built to survive a public crowd, all day, every day.

role · real-time systems & AI integration stack · TouchDesigner · Gemini API
View Project →
AI Chef System
multi-model AI · NLP

AI Chef System

A multi-model pipeline that generates complete, personalized recipes — the right dish for the right person, with imagery to match. LLMs handle the cooking logic; image models plate it.

role · pipeline design & prompt systems stack · LLMs · image generation
View Project →
AI Photobooth
stable diffusion · real-time

Real-time AI Photobooth

One of the first live Stable Diffusion photobooths — before this was a product category. Walk in, get photographed, walk out as generative art. Latency tuned until the magic felt instant.

role · concept & full build stack · Stable Diffusion · TouchDesigner
View Project →
Burger King AI Pipeline
content pipeline · generative AI

Burger King AI Content Pipeline

End-to-end generative pipeline for a global campaign — LLMs and image generation working inside brand guardrails, producing on-brand content at a scale a human team couldn't match.

role · pipeline architecture stack · LLMs · image generation · automation
View Project →
earlier work · 4× FWA awarded
02 / earlier work

Creative Technology

Interactive work spanning AR, physical computing, accessible design, and brand installations.

Pharmaton Kids Website
FWA · accessibility · web

Pharmaton Kids — Accessible Web

Inclusive design for children with reading disabilities — adaptive UI, simplified navigation, and accessible interactions.

FWA Case →
Parkinson's Campaign
FWA · 3D printing · health

Printed by Parkinson's

Tremor data translated into 3D-printed objects — making an invisible condition tangible.

FWA Case →
🎄
FWA · AR · volumetric video

AR Christmas Card

Web-based volumetric video experience — a holiday message you could walk around.

FWA Case →
👟
FWA · installation · motion

Adidas Originals Launch

Motion-sensor installation for the Deerupt launch — real-time graphics driven by how you move.

FWA Case →
03 / the human

About

I'm Luis — theguaz online: Guajardo + Díaz, a handle I coined back when Hotmail was the new thing, and never let go. A Chilean creative technologist based in Hilversum, working at .Monks. I started in interactive websites and physical installations over fifteen years ago, and moved into generative AI early — building one of the first real-time Stable Diffusion photobooths back when that required duct tape and conviction.

My favorite brief is the one without a reference. I prototype fast, mix physical computing with AI models, and obsess over the last 10% — the part people actually feel.

Also: my cat Canela is now a Walmart Chile brand character. Possibly my proudest production deploy to date.

base
Hilversum, NL · CET
experience
15+ years
recognition
4× FWA awarded
currently
Senior Creative Technologist @ .Monks
availability
Open for freelance · remote worldwide
let's build something

Got an AI project?
Let's talk.

I work with brands, agencies, and startups to build AI-powered experiences that are actually interesting. From proof-of-concept to production pipeline — first call is always free.

[email protected] @theguaz Hilversum, NL · working worldwide