● 4 diagnostic slots left / October 2026

I build automation systems for marketing agencies and e-commerce brands.

17 years building production systems for companies like ENI ($101B), TIM ($16B, a 5M-user product) and Decentraland. Then I automated my own e-commerce brand, traffic to reporting, and sold it. You hire me, not a team.

Book a free fit call →See Example Systems

15 min, no deck, no pitch. We check whether your workflows are worth automating.

[ Fig. 1 ]

MANUAL REPORTING → KILLED ·LEAD FOLLOW-UP → AUTOMATED ·CRM UPDATES → INSTANT ·CLIENT ONBOARDING → ONE FORM ·INTERNAL AGENTS → BOOK APPOINTMENTS AT 2AM ·ENI · DECENTRALAND · TIM · AGÜITA ·MAKE · N8N · PYTHON · GA4 ·NO ZAPIER SPAGHETTI ·CLAUDE CERTIFIED ARCHITECT · BY ANTHROPIC ·

THE FOUNDER

One engineer. Same email thread, end to end.

I'm Fabio. 17 years building and running production systems before The Automate Lab. Backend services, simulations, dashboards. As a contracted engineer: ENI ($101.4B rev), Decentraland ($51M raised), TIM ($16B rev). Then I acquired, ran, and sold Agüita, a stainless-steel bottle brand in Tenerife — imports from China suppliers, a real warehouse, in-house laser engraving, B2B wholesale to hotels and golf courses. The first test bench for these systems. Now I build them for marketing agencies and e-commerce brands.

Most "AI automation" agencies are one Zapier flow away from breaking. I build the way engineers build: error handling, monitoring, version control, rollback. A workflow that gives you your week back = engineering. A workflow that costs you a week debugging = duct tape. Anthropic certified me as a Claude Certified Architect (Foundations) in September 2026. The exam covers exactly this: agents that finish the job or hand it to a person, with a record of what happened.

You hire me directly. No account manager, no juniors, no handoff to anyone. Whoever scopes your Sprint ships the system. Same person, same email thread, end to end.

Claude Certified Architect - Foundations

Issued by Anthropic, September 2026

Verify on Credly

[ FIG. 2 ]

WORKS WITH YOUR STACK

If your stack is here, it can be automated.

OUTCOMES

What This Changes

1,600

daily organic visits, up from 200 in 5 months (Agüita, AI content pipeline: 8–9 posts a month, review-only)

~10h

/month of owner reporting killed. Monday numbers land in Slack at 7am, untouched (Agüita)

2→9

recurring clients for my own drone-video side business, from cold email to realtors.

1 form

replaces 6 onboarding steps. one form.

15h

/week of data work killed at TIM ($16B rev)

Your numbers depend on your workflows. that's what the Sprint is for.

/ WHAT I BUILD

[ FIG. 3 ]

What I actually build.

I map your manual workflows, kill the ones that shouldn't exist, and replace the rest with systems that run without you. Same approach I used to kill 15 hours/week of data work at TIM.

Lead capture, routing, and follow-up automation
Client onboarding flows that run on a single form
Reporting dashboards that run themselves
One founder, full-stack execution, no handoffs

CAPABILITIES

What I Build

[ FIG. 4 ]

Lead Capture + Follow-Up

→ Stop losing the 1 to 3 inbound leads a week that ghost because nobody replied in time. answered in under 5 minutes, every time.

Maken8nHubSpotSlack

[ FIG. 5 ]

Automated Reporting

→ Your Monday numbers in your inbox at 7am. 6 hours/week back to the founder.

GA4PythonSheetsMake

[ FIG. 6 ]

Client Onboarding

→ Onboard a new client in 10 minutes. one form fires contract, invoice, welcome, workspace, channels. all six.

TypeformMakeNotionStripe

[ FIG. 7 ]

AI Receptionist (and other internal agents)

→ Books appointments at 2am. answers your top 5 questions in your voice. never sleeps.

n8nMakeZapierAI Agents

[ FIG. 8 ]

Landing Pages & Web Systems

→ When your funnel needs a page, I build the page. not the agency thing.

Next.jsTailwindPythonAstro

CASE STUDIES

Four systems, four manual jobs killed.

[ FIG. 9 ]

CASE-LOG / 01

All4cycling automated its 80/20 strategy across 100k+ variants and saves 10 hours of work a month

Problem: All4cycling pulled the ABCDE classification of its Shopify variants by hand, from Mipler and Excel. 30 minutes every 3 days.

System: Every night, an automation classifies the active variants from A to E on the last 180 days of sales, writes the class back to Shopify and emails the full report the next morning.

Tools: Shopify Admin APIGitHub ActionsPythonResend

Result: 10 hours a month back to the marketing team. The classes are fresh every morning instead of every 3 days.

★★★★★

The project required sharp logic, seamless Automates Shopify API Integration , and reliable daily scheduling—and Fabio delivered on every front. The logic for filtering revenue tiers (80/15/5/0) across categories was implemented flawlessly, and the automated daily cron job runs like clockwork. Communication was clear, deadlines were met ahead of time, and the code is clean and scalable. If you need a skilled Automation Engineer who deeply understands e-commerce data, Shopify APIs, and automation, look no further. Would definitely hire again!

Marco, All4cycling · Upwork review

[ FIG. 10 ]

CASE-LOG / 02

Agüita / the owner's reporting job, killed

Problem: Agüita, the stainless-steel bottle brand I acquired, ran, and later sold in Tenerife (physical stock, B2B wholesale, in-house laser engraving), had me as its reporting department. Every week I pulled Shopify and GA4 numbers by hand to see what was selling and where the store stood: two to three hours of it, every week. First job I fired myself from.

System: Make scenario pulls Shopify + GA4 data nightly → writes to a Google Sheets master → digest lands in Slack every morning, summary every Monday at 7am. Error alerts on failure.

Tools: MakeGA4SheetsSlack

Result: Reporting ran itself from launch until the day I sold the brand. The same brand also ran an automated revenue-split report for a collab: tagged-SKU sales, extracted and delivered to the partner, another three to four hours a month I stopped spending. I was the test bench.

[ FIG. 11 ]

CASE-LOG / 03

TIM ($16B rev) / 15 hours/week of data work, gone

Problem: While I was a contracted engineer inside TIM ($16B revenue), my team was burning 15 hours a week on the same data-crunch: pulling, cleaning, reconciling, exporting. Manual every Monday.

System: Scheduled Bash cron → Python scripts pull from source systems → clean, reconcile, validate → write outputs in the same downstream format. Logs to disk. Failure alerts before anyone notices.

Tools: PythonBashCronCSV/Excel

Result: 15 hours/week removed. ~60 hours/month of senior analyst time recovered. Monday bottleneck eliminated.

[ FIG. 12 ]

CASE-LOG / 04

My drone-video service / 2 to 9 recurring clients

Problem: My own drone-video side business (real-estate videos, commercials, documentaries) had 2 recurring clients, all from word of mouth. I was flying shoots all day, zero time left for prospecting. So I became my own first lead-gen client.

System: Apollo scrapes and filters real-estate agencies → OpenAI writes a personalized opening line per prospect → Mailgun sends the sequence with reply detection → replies land in Pipedrive, tagged and ready.

Tools: ApolloOpenAIMailgunPipedrive

Result: 2 to 9 recurring clients total, from cold email to realtors. My only job left in the loop: answering warm replies.

Four real systems. All4cycling is a client project, and Marco approved the text and the numbers. The other three ran in businesses I owned (a bottle brand I later sold, my own drone video service) and under contract inside TIM. No invented results.

THE BOTTLENECK SPRINT

Start with a paid Bottleneck Sprint.

Before I quote a project, I map the one workflow that costs you the most. You get a one-page breakdown: how it runs today, the hours and dollars it burns every month, and a straight yes/no on whether automating it pays back inside 90 days.

What you get:

→A one-page map of your highest-value workflow.
→The measured time and cost of doing it manually, per month.
→Yes/no on 90-day payback, with the number behind it.

Credited in full against your project if you start within 30 days.

Book a free fit call →

Book the free fit call first. If we're a fit, the Sprint is the next step.

PROCESS

How The Automate Lab Works

Diagnose

I map your current systems and every manual process eating your week.

Week 1

I map your current systems and every manual process eating your week.

Design

I architect the automation blueprint: what gets automated, integrated, and cut.

Week 2

I architect the automation blueprint: what gets automated, integrated, and cut.

Build

I build and deploy your systems. you review, I refine, it goes live.

Weeks 3–4

I build and deploy your systems. you review, I refine, it goes live.

Launch

Systems go live and I monitor performance, optimising what matters.

Ongoing

Systems go live and I monitor performance, optimising what matters.

BELIEFS

Most agencies sell you complexity. I sell you back your week.

Automate it or stop complaining about itSystems > heroicsEngineering, not duct tapeReplace the stack, don't add to it

FAQ

Common questions.

What if I want to maintain this myself later?

Every system ships with documentation, version-controlled code, and a 30-min handoff session. If you have a developer in-house, they can take it over. If not, retainer is available.

How long until I see results?

Most lead-routing and reporting systems are live within 2–3 weeks of the Sprint. Larger onboarding or multi-stack integrations are 4–8 weeks. The first measurable outcome appears in the first week after launch.

What if my stack isn't on your list?

The 10 tools listed are common. I work with anything that has an API or a webhook. Mention your stack on the fit call; if it's not feasible, I'll tell you on the spot.

Do you do retainer / ongoing optimization?

Yes. After your first system is delivered, retainer covers ongoing monitoring, optimization, and new automation work. Month-to-month, no annual commitment.

Will the call be a sales pitch?

No. The free fit call is 15 minutes to see whether your workflows are worth automating. If they are, the paid Bottleneck Sprint is the real diagnostic, and you leave with a written deliverable either way. No follow-up sequence.

READY?

Book a free fit call.

Fifteen minutes to see if your workflows are worth automating. If they are, the next step is a paid Bottleneck Sprint: a one-page map, the cost of your manual work, and a yes/no on 90-day payback.

Book a free fit call →

15 minutes, real workflow, real numbers. No deck, no pitch.

Bottleneck Sprint, credited in full against your project.

OPERATIONAL

[ Fig. 14 ]