The worst scenario is when everything seems to work.
Your tracking isn't visibly broken. But underneath: inconsistent definitions, vendor-shaped schemas, models trained on incoherent signals.
You're measuring noise efficiently. I design measurement systems that define what matters before tools execute anything. Architecture first. Implementation second.
I design the architecture. I do not sell daily tracking implementation.
Start a Strategic ConversationWhy This Matters Now
"AI is about to scale every decision you make. If your data definitions are ambiguous, you aren't just measuring noise—you are training your systems to amplify it. Architecture is no longer a 'nice to have' optimization; it is the prerequisite for automation."
01 / The Context
You recognize yourself here...
Your data team keeps 'fixing' tracking, but six months later it's broken again.
Different teams get different numbers from the same events, and nobody knows which one is right.
You're preparing for AI and realize your definitions aren't clear enough to automate.
Every tool change breaks everything downstream because your schema memorized the vendor.
Your definitions of 'customer' vary by department, making cross-functional analysis impossible.
You want to move from 'this works enough' to 'this is designed to scale'.
02 / The Approach
Architecture First
The quality of data is defined before the first line of code is written.
I decide what gets measured, how it is named, and what rules it must follow.
I do not compete with AI on execution speed. I provide the judgment that AI lacks.
I deliver the clear rules that allow your internal teams, vendors, or AI agents to execute without ambiguity.
1. Definition Layer
Events, Identities, Metrics. What an event actually means when someone looks at a dashboard.
2. Constraints Layer
Privacy, Consent, Business Logic. What data is allowed to exist, even if it would be useful.
3. Execution Interface
The spec where tools, AI, and vendors operate without making interpretation decisions.
"Automation creates speed. Only architecture creates direction."
03 / Services
How I Help
Concrete measurement systems designed to survive tools, vendors, and AI-driven execution.
Low risk. High clarity.
1. Strategic Diagnostic & Roadmap
A rapid audit of your measurement architecture to identify what's broken and what order to fix it in. No generic recommendations — just a clear analysis of where your foundations are weak. Outcome: Certainty about what's actually wrong (not just symptoms) and a prioritized roadmap.
- →Uncertainty about where to start
- →Debates about tool vs. architecture problems
- →Confusion about symptoms vs. root causes
- →Risk of investing in the wrong solution first
- ●Current State Architecture Map
- ●Gap Analysis Report
- ●Prioritized Remediation Roadmap
- ●Tool vs Architecture Assessment
The foundation for everything else.
2. Measurement Architecture Design
I define what should be measured and what it means in your business — before implementation begins. Outcome: A complete architectural specification that implementation teams execution without interpretation decisions.
- →Ambiguity in data definitions
- →Broken tracking across platforms
- →Lack of ownership in data collection
- →Events that exist but nobody knows what they mean
- ●Event architecture & business logic mapping
- ●Canonical definitions framework
- ●Cross-platform identity design
- ●Measurement governance model
- ●Execution specification
- ●Validation & testing criteria
Builds on top of the core architecture.
3. DWH-First Measurement Systems
I design systems where your warehouse owns your data — vendors just consume it. Outcome: Your measurement architecture survives vendor changes. When you switch tools, you change a transformer, not your entire data model.
- →Vendor lock-in (GA4, Mixpanel, etc.)
- →Conflicting numbers across tools
- →Fragile tracking that breaks on tool changes
- →Loss of historical data during migrations
- ●DWH Ingestion Logic Specification
- ●Canonical Data Model Design
- ●Downstream Feed Architecture
- ●Identity Resolution Framework
- ●Vendor Transformation Logic
Don't let legal kill your data.
4. Consent & Measurement Architecture
I bridge the gap between privacy compliance and measurement utility — so you can respect consent without going blind. Outcome: A clear technical specification for how consent controls what data flows where.
- →Fear of GDPR/ePrivacy fines
- →Data loss due to overly aggressive blocking
- →Implementations that break when regulations update
- →Misconfigured CMPs
- ●Consent Architecture Specification
- ●CMP Integration Design
- ●Data Flow Governance Rules
- ●Anonymization & Pseudonymization Logic
- ●Vendor Data Sharing Controls
I don’t sell implementation.
I design the system that decides what should be implemented — by teams, vendors, or AI.
04 / Why I Do This
The Insight
I've spent the last decade in the trenches of data collection. I've seen millions of euros wasted on tools that promised to 'democratize data' but only democratized confusion.
My philosophy is simple: The problem is rarely the tool. The problem is that nobody designed the data model. They just started tagging things.
I built this practice to help companies who are tired of the cycle of 'implement, break, fix, repeat'. I want to help you build a system that lasts.
05 / Authority
Strategic rigor. Technical depth.
Head of Data Collection at Fever
System Design
+7 Years designing high-volume architectures. I build systems that handle complexity so teams can handle decisions.
View case studies →Education
Professor of Data Analytics at EDEM & MIOTI. I teach the principles of measurement to the next generation of leaders.
Standards
Author of official templates in the Google Tag Manager Gallery. I contribute to the open standards that power the industry.
06 / Track Record
Proven Architectures
Billions of events. DWH-First.
Designing a DWH-first collection system for a global platform processing billions of events.
Privacy & Governance
Integración de OneTrust y mParticle. Definiendo el gobierno para un flujo de datos privacy-centric.
Multi-tenant Consent
Designed the consent implementation architecture for all group companies.
07 / Fit Check
This is NOT for you if...
- ✕You are looking for a 'hands-on' developer to close JIRA tickets.
- ✕You value tool configuration over system definition.
- ✕You believe AI can 'just figure out' your business logic without guidance.
- ✕You want a monthly retainer for operational maintenance.
- ✕You prefer speed of implementation over reliability of data.
- ✕You view data collection as an IT task, not a strategic asset.
If any of these describe you, that's okay. I'm simply not the right partner.
08 / Contact
Start a Strategic Conversation
If any of the situations above sound familiar, this is usually the right place to start. Explain your context. I read every message and reply if I can add value.