Systems that ship
Local AI Coding Agent: a 24-task evaluation corpus. Website Audit Agent: evidence before synthesis. DemandOS: deterministic forecasting on synthetic data. IRIS: public orchestration architecture with synthetic examples.
RAÚLMERMANS
I build across AI, business, product, and creative work.
Independent systems are the centre of my practice. Marketing, brand, CRM, and visual work give me the context to build tools that answer real problems.
Commercial systems
Lifecycle, campaigns, segmentation, and adoption work that made systems answer to real operational constraints.
Taste as interface
Photography, visual studies, luxury references, and brand systems used as the judgment layer around technical tools.
About
A profile built from work, place, and practice.
Operational taste means knowing what to structure, what to leave editable, and when people need to decide.
Málaga is the origin. Madrid sharpened the practice through IE University, marketing, CRM, and commercial teams.
Orlando changed how I read service. Commercial systems added the pressure of lifecycle work, segmentation, partners, and scale.
The portfolio turns that thinking into campaign tools, data products, brand systems, and visual studies.
Operating profile
- BaseMadrid
- OriginMálaga
- EducationBBA Marketing, IE University
- Working contextCRM, lifecycle, partner programmes, AI adoption
- Independent workCampaign tools, data products, brand workflows, internal systems
- Visual practicePhotography, album covers, image systems, art direction
- AI systemsLocal models, Ollama, MCP, tool calling, evaluation
- EngineeringTypeScript, Python, React, Next.js, SQL, PostgreSQL
- Data & automationForecasting, reporting pipelines, n8n
Geography3 cities, 5 languages
Málaga36.7N · 4.4Whome
Madrid40.4N · 3.7Wnow
Orlando28.5N · 81.4Wexchange
Languages
- ESSpanishNative
- ENEnglishNative
- NLDutchBasic
- DEGermanBasic
- ITItalianBasic
Case Studies
Apps
Visuals
Photography
Chapters, not job titles.
A compact read of the contexts that shaped the work: markets, delivery, service, building, scale, and the current loop.
01
Madrid — learning the language of markets
IE University · BBA Marketing
Business, marketing, digital analytics, and the first technical vocabulary for turning customer behavior into decisions.
MadridIE UniversityMarketing
02
Teams — learning delivery
IE Marketing Lab
Project work with deadlines, partners, and team handoffs: strategy only counted when it became something delivered.
LeadershipStrategyDelivery
03
Orlando — seeing service differently
UCF exchange
A semester at the University of Central Florida sharpened the read on service, retail behavior, and cultural expectations.
OrlandoUCFService
04
Independent practice — learning to build
Portfolio systems
Campaign tools, data products, image systems, and site architecture turned positioning into usable interfaces.
PrototypesCase studiesVisuals
05
Commercial systems — working inside scale
CRM · lifecycle · partner programmes
Large audiences, commercial cadence, segmentation, partner expectations, and internal AI adoption make the work accountable.
CRMLifecycleScale
06
Now — building the loop
Tools · brand workflows · visual practice
Building the loop between workflow logic, interface design, brand memory, data boundaries, and human review.
SystemsInterfaceTaste
How I work.
The work starts with the decision and ends with a surface people can review, use, and improve.
Decision
Start with the decision.
What does the user need to understand, choose, approve, or repeat?
QuestionContextCriteria
Workflow
Map the workflow.
Inputs, constraints, handoffs, review points, risks, and outputs.
InputsHandoffsRisks
Interface
Build the interface.
Prototype the surface where the work becomes visible and usable.
SurfaceStatesFeedback
Intelligence
Add intelligence carefully.
Use AI or ML where it improves speed, synthesis, or diagnosis — not where it removes accountability.
AIMLReview
Taste
Keep taste in the loop.
Brand voice, image logic, pacing, and visual judgment remain part of the system.
VoiceImage logicPacing
Let's make something work.
Open to roles, projects, and collaborations
raulmermans@gmail.com →View technical work on GitHub ↗Madrid · Remote · EU