Àlex Català EN / ES
METHOD
HOW I WORK

Marketing, communication and design — and AI to build the systems that take them far beyond the ordinary.

My base is creative: marketing, branding, advertising, design, communication. AI isn't the goal — it's how I work faster and build systems that reach a level the ordinary can't. I encode the craft once, let the system run, and keep a human hand on everything that has to feel human. Less hype, more proof: not slides, shipped work.

SYS

01 Systems

I build the systems: brand understood, rules in a CLAUDE.md, a pipeline that runs only where I point it.

See the systems →
DG

Diseño gráfico y vídeo — Claude Design + Higgsfield

En la parte visual diseño con Claude Design y Higgsfield. Con Higgsfield puedo producir vídeos visuales, campañas publicitarias y casi cualquier cosa — el mismo método de control, aplicado a la imagen y el movimiento: la marca primero, la IA solo rellena el carril que marco.

Claude DesignHiggsfieldVídeoCampañas publicitarias

He aprendido estas herramientas a buen nivel — pero todavía no tengo un proyecto terminado que mostrar aquí. Pronto.

01 — 05
THE METHOD · HOW I ACTUALLY WORK
01 Understand the brand
First I understand the brand exactly where it is — its voice, its rules, its client. Nothing starts until I've understood it myself.
02 Order it for the AI
I order everything so the AI can understand it. I work the way I was trained — at university and in practice; the AI adapts to my judgment, never the other way around.
03 Rules · CLAUDE.md
I set the rules in a CLAUDE.md — what's allowed and what isn't. That's the base that stops the AI from doing whatever it wants.
04 Pipeline + workflow
On that base I build a pipeline with its workflow, driven by careful prompting — everything ordered inside the CLAUDE.md.
05 Steer & build
With commands, tools and agents I steer the AI exactly where I want — then build the projects, which can branch in different directions.
VAULT
HOW I MAKE A PIPELINE · STEP BY STEP
📁 01 Proyectos
Bibì e Bibò.md
leParfum.ai · ICM.md
Inversiones.md
📁 04 Catálogo
bibi-e-bibo (buzón).md
Swedish Tonic.md
📁 05 Contenido
site-offline · copy.md
video-pipeline · playbook.md
📁 _Buzón
⤓ drop new work here
ClaudeBriefBrand voicePipelineEvalsProductPostImageReelDashboardDecisionRaw information
01

Collect everything

First, a big information dump — sources, references, screenshots, notes. Raw and messy, no order yet.

02

A folder system the AI can read

I drop it all into a structured folder system: clear names, one idea per file — built so Claude can actually understand it.

03

Claude reads + converts

Claude goes through the folders and converts them into clean, structured notes — explained in a way it can reuse later.

04

The Obsidian map

We build the Obsidian maps. Now the information is connected, colour-coded by zone and far better organised — for me and for the AI. It stays updated: a constant, regulated flow, every zone attended.

05

Any proposal runs

From this organised brain the AI can make any proposal and the pipeline runs end-to-end — caring for each point, each zone. That's how I use it.

SYSTEMS
THE SYSTEMS · THE EXACT PIPELINE

Brand understood and ordered, rules in a CLAUDE.md, a pipeline that only goes where I point it — here are the exact systems I've built.

01

Bibì e Bibò — invisible-AI content system

A gelato brand with seven locations can't hire a writer per city, and one central voice reads flat. So I encode the brand voice once, then a Claude pipeline drafts the month's strategy, calendar and copy — re-toned per city with its own festivities and flavour of the week. Higgsfield evokes the real origin of each ingredient. Nothing publishes until it clears a 10-point anti-detection checklist: if it reads as AI, it doesn't ship. The owner approves in minutes; it runs in ~13 h/month at ~€50/month of tooling.

Claude APIPythonHiggsfieldAnthropic SDK
Monthly briefing per location (flavour of the week, local events, promos)
Brand voice + hard rules, encoded once (tone, do/don't, banned words)
Claude drafts strategy + monthly calendar
Claude writes copy + captions, re-toned per city
Higgsfield generates ingredient-origin visuals
10-point anti-detection check · human approves
Schedule + publish · 7 Google profiles + social
02

leParfum.ai — fragrance image pipeline

An AI-native fragrance brand needs evocative product imagery at scale — without the tell-tale "AI stock" look. I built a taxonomy that maps fragrance notes to image moods, then a prompt library tuned to the brand's vocabulary. Gemini personalises and Higgsfield adds the cinematic layer, but every output obeys inviolable bottle rules: the bottle leads, and no URL or QR is ever baked in — text and logo are added later, in-brand. Outputs flow straight into the brand grid.

GeminiHiggsfieldComfyUIFramer
Fragrance notes + mood brief
Map notes → mood taxonomy (each note → a visual register)
Brand-tuned prompt library
Gemini · personalised base image
Higgsfield · cinematic layer + motion
Inviolable bottle-rules check (bottle leads · no URL/QR baked in)
Add text + logo later, in-brand
Into the brand grid
03

Inversiones — geopolitical intelligence system

Investment signals are scattered across dozens of feeds with nowhere to collect, score and remember them. I built a FastAPI backend that ingests GDELT, Google News across 12 regions, RSS and CoinGecko, geolocates each event on a world map, and scores its market impact with a rules-based engine — no API key, ~0 cost. Everything persists in SQLite and surfaces in a React + Leaflet dashboard, while an Obsidian vault is the system's long-term brain: signals → analysis → decisions.

FastAPIReactSQLiteGDELT
Sources · GDELT · News (12 regions) · RSS · CoinGecko
FastAPI collectors · scheduled pulls
Dedupe + geolocate on a world map
Rules-based impact score (no API key · ~0 cost)
Store in SQLite
React + Leaflet dashboard
Obsidian vault · long-term memory
Signals → analysis → decisions
04

Computer Use · MCP + Premiere — my own tools

Most desktop apps have no API — so Claude can't touch them. I build the bridge. A Node MCP server (single index.js, stdio) exposes 14 tools that let Claude see and operate macOS: screenshots, files, Terminal, arbitrary AppleScript. A second bridge talks to Premiere over file-IPC (8 tools, ExtendScript) so Claude can cut and assemble 9:16 reels from AI clips. These are the "arm and eyes" that make every other system run — local-first.

Node.jsMCP SDKAppleScriptExtendScript
Claude receives a request
MCP server · stdio · 14 tools
macOS · Finder · Terminal · AppleScript
Claude composes the multi-step op
Premiere bridge · file-IPC · 8 tools
ExtendScript · cut + assemble timeline
Export 9:16 reel
DECISION
HOW I BUILD A SYSTEM — SAME KIT, DIFFERENT SYSTEMS
Brief + brand voice · one page
Recurring on-brand content? → Content pipeline (e.g. Bibì e Bibò)
Visuals or video at scale? → Image / video pipeline (e.g. leParfum.ai)
Scattered signals to track? → Intelligence dashboard (e.g. Inversiones)
An app Claude can't reach? → Build an MCP tool (e.g. Computer Use)
QUICK LINKS
Àlex Català
©2026 · Independent practice · Barcelona