Maximum flexibility, working with your stack
I work inside your platforms and your data. I don't sell proprietary tooling: I build something ad-hoc based on your business needs.
bolt Powered by AI · MCP-native
Boutique Marketing Engineer for your growth strategy
A holistic analysis of your business, from acquisition to retention, built on engineering infrastructure: instrumentation, automation, and AI agents wired to your platforms with the right context. One person operating as a complete team.
verified Recommended by Visionarios graph_3 5+ active accounts hub Ad-hoc connections
Reading warehouse · running model
Your stack · Always on
Clients already operating with a Marketing Engineer
The Marketing Engineer process
Four layers. One system.
From campaign setup to the full business report. Each layer is built on top of the previous one, and each is documented and versioned as code with full transparency.
Measure right before optimizing
I make sure all measurement is correctly implemented. I connect data from the CRM and/or ERP with acquisition channels and reports. Without clean data there are no clean decisions.
Tracking audited Conversions that matter CRM connected
GTM · container GTM-9X4P2N
Active tags
ads_clickGoogle Ads: Conv. Leadlive
campaignMeta: CAPIlive
analyticsGA4: purchaselive
linkLinkedIn Insightpaused
cloud_syncServer-side · sGTMlive
// dataLayer.push: lead_qualified event: "lead_qualified", deal_value: 12400, deal_currency: "EUR", source: "google_ads", campaign_id: "22041883", consent_state: "granted", hubspot_deal_id: "118-9302"
Campaigns that learn from your business, not the pixel
Account structures optimized for quality signals: HubSpot leads, real closed-deal values, segmentation based on business data. Every technical and attribution pain point of the marketing team, covered.
Offline conversions Value-based bidding Cross-channel
Google Ads · MCC overview
| Campaign | Spend | Conv. | Deal value | tCAC |
|---|---|---|---|---|
| PMax · ABM Tier-1value-based bidding | €4 200 | 23 | €78 400 | €700 |
| Search · Brand defenseexact match | €820 | 18 | €42 300 | €312 |
| Demand Gen · LookalikeHubSpot LTV-90 audience | €1 980 | 11 | €26 100 | €640 |
| LinkedIn · ABM seed list500 Tier-1 accounts | €2 500 | 9 | €32 800 | €625 |
east HubSpot deal closed-won → offline conversion → bidding learns from real revenue, not the lead
Agents that analyze your account every morning
I add an AI layer to your platforms via custom MCP servers. Every morning you receive in your chat: spend vs target, anomalies, scaling opportunities, alerts, and a link to the report with up-to-date data.
Custom MCP Slack daily digest Anomaly detection
Slack · #growth-daily
ME
Daily digest · Mar 12
Spend on track (€892 / €900 plan). 2 anomalies detected. 1 scaling opportunity.
Spend on track (€892 / €900 plan). 2 anomalies detected. 1 scaling opportunity.
Spend 7d€6 240 −4% vs target
Blended CPL€42.10 −18% MoM
PMax · Tier-1+38% conv. scale candidate
Meta · RetargetCTR 0.4% creative fatigue
# mcp-server-google-ads · tool definition
@tool("get_campaign_anomalies")
def detect(account_id: str, window: str) -> list:
"""Returns campaigns with z-score > 2.0 on CPL."""
return client.query("SELECT ...")
@tool("get_campaign_anomalies")
def detect(account_id: str, window: str) -> list:
"""Returns campaigns with z-score > 2.0 on CPL."""
return client.query("SELECT ...")
Reports that everyone involved understands
Weekly and monthly reports with real CPL (not platform attribution), impacted pipeline, and projections. Auto-generated using the source of truth: your CRM crossed with your acquisition channels.
Real CPL Attributed pipeline Auto-generated
Report · Week 11
Real CPL (post-CRM) ▼ 18%
€42.10
| Channel | Leads | SQL | Pipeline € | Real CPL |
|---|---|---|---|---|
| Google Ads | 87 | 23 | €78 400 | €48 |
| Meta | 41 | 8 | €18 200 | €44 |
| LinkedIn ABM | 19 | 11 | €32 800 | €131 |
| Organic + SEO | 62 | 14 | €41 200 | n/a |
Core capabilities
From data to action, in a single platform
I work inside your kitchen and orchestrate it with MCP, AI, and a lot of context. These are the six capabilities I deliver in every project I work on.
If your tool has an API, it can be integrated
Custom MCP servers that connect AI to your stack: ads, CRM, analytics, automation, and messaging in a single language.
Google Ads
Meta Ads
HubSpot
GA4
LinkedIn
GTM
Slack
Notion
Looker Studio
Search Console
Make
Zapier
Shopify
Klaviyo
Brevo
Gmail
TikTok Ads
YouTube AdsFrom click to closed deal
End-to-end attribution with disciplined UTMs, CRM conversions, and validation against the CRM. You know which campaign closes each deal.
| Source | Spend | Leads | SQL | Won | CAC |
|---|---|---|---|---|---|
| Google Ads | €4 200 | 87 | 23 | 6 | €700 |
| Meta | €1 800 | 41 | 8 | 2 | €900 |
| €2 500 | 19 | 11 | 4 | €625 |
Daily digest in Slack
Every morning, an agent analyzes your accounts and posts the summary in your channel with a link to the report. Zero friction, we all speak the same language with the same data.
ME
Spend 7d€6 240 −4%
Blended CPL€42.10 −18%
PMax Tier-1+38% scale
Documentation and handover
Every account is documented in Github: tracking architecture, naming conventions, playbooks. Your team and the AI can keep operating without me.
folder Growth Operations
article Tracking architecture
code_blocks dataLayer spec · v3.2
terminal Server-side runbook
article Naming conventions
article Playbooks
play_arrow Scaling PMax
I work like an engineer
Versioning, monitoring, debugging, and experiments with explicit hypotheses. With automation, growing a company turns into a method.
EXP-024 PMax with HubSpot LTV-90 audience will reduce CAC in Tier-1 vs standard lookalike −28% CAC SHIP
EXP-023 Demand Gen creative-led will beat Search in lead volume while keeping CPL +34% CPL KILL
EXP-022 Server-side conversions with real deal_value will improve bidding signal +19% conv. SHIP
EXP-021 LinkedIn ABM seed-list will generate SQLs with cycle < 30d ±2% LEARN
The secret no one tells
Context: the key to success with AI.
The most expensive mistake when adopting AI in marketing isn't picking the wrong model, nor the most complex connector. It's giving the model a business it doesn't understand at all. The winner isn't whoever has more AI, it's whoever knows best how to provide the business context.
ChatGPT connected via MCP to your Google Ads doesn't know you sell B2B SaaS with sales cycles longer than 90 days. It doesn't know your budget sits around €10,000. It also has no idea that Brand Search campaigns don't deliver incremental conversions, only defense. It doesn't know the main channel is Paid but the one that converts best is Webinar. Without the right context, it gives plausible, generic answers. With the right context, it gives decisions.
That's why the first deliverable of every engagement isn't an MCP server or an agent. It's a structured discovery phase: guided surveys, interviews with sales, mapping of your real funnel, internal company vocabulary. All of that is synthesized and given to the model as persistent context, before touching a single integration.
With business knowledge plus technical marketing expertise, AI becomes a machine for making the right decisions.
Anyone can implement a model. Defining the context well is what gets paid for.
The method, audited
Marketing engineering in numbers
0+
Active accounts with constant analysis via MCP agents
0
Integrated platforms via custom MCP servers
0%
Average reduction in cost per conversion within the first 180 days
The method shift
Traditional marketing vs Marketing Engineer
Same discipline, an extra layer of thinking. One operates with opinion and manual reports; the other with systems, hypotheses, and crossed data.
Traditional marketing
Manual reports and pixel attribution
- closeManual reports in PowerPoint or Excel every Monday that can break
- closeUsually pixel attribution, without crossing against real data
- closeOptimization based on impulses
- closeFragmented stack, each platform in its own silo
- closeSells you hours of work, not systems
Marketing Engineer
Systems, methods, and crossed data
- checkAutomatic daily digest where we all speak the same language
- checkCross attribution with CRM: real CPL, not the pixel's
- checkReports and ad-hoc tools for your business
- checkStack orchestrated via MCP, one language for every platform
- checkDelivers systems that scale, not hours that get consumed
Ready to operate your growth as a system?
Free 30-minute diagnostic: I review your stack, strategy, and current method. Tell me your situation and I'll get back to you in less than 24h.
- check_circle Tracking and attribution audit
- check_circle Stack diagnostic and MCP opportunities
- check_circle Roadmap for the first 20% improvement