You will get an AI agent that runs your marketing busywork on autopilot
Rising Talent
Rising Talent
Project details
Your team is doing work a reliable agent should handle: qualifying leads, drafting replies, enriching CRM records, moving data between tools, summarizing calls, tagging tickets and assembling recurring reports.
I build production AI workflows in n8n or Make, powered by the model that fits the task and budget. Common builds include lead enrichment and scoring, inbound-email triage with one-click approval, content pipelines with brand checks, meeting summaries written into HubSpot or GoHighLevel, and internal operations that sync systems automatically.
The difference is operational discipline: retries and failure alerts, human approval for customer-facing actions, logging, documentation, a maintenance plan and token-cost controls. Starter covers one defined integration across two tools. Standard delivers one AI workflow across up to three tools. Advanced delivers a multi-step agent with decisions, tool calls, up to six integrations, human gates and two weeks of tuning.
Your accounts and API keys remain yours. Larger systems with connected agents, dashboards and team training start as a custom scope after a short process review.
I build production AI workflows in n8n or Make, powered by the model that fits the task and budget. Common builds include lead enrichment and scoring, inbound-email triage with one-click approval, content pipelines with brand checks, meeting summaries written into HubSpot or GoHighLevel, and internal operations that sync systems automatically.
The difference is operational discipline: retries and failure alerts, human approval for customer-facing actions, logging, documentation, a maintenance plan and token-cost controls. Starter covers one defined integration across two tools. Standard delivers one AI workflow across up to three tools. Advanced delivers a multi-step agent with decisions, tool calls, up to six integrations, human gates and two weeks of tuning.
Your accounts and API keys remain yours. Larger systems with connected agents, dashboards and team training start as a custom scope after a short process review.
AI Algorithms
Large Language ModelAI Applications
Conversational AI, Natural Language Generation, Natural Language UnderstandingAI Models
ChatGPTWhat's included
| Service Tiers |
Starter
$197
|
Standard
$497
|
Advanced
$897
|
|---|---|---|---|
| Delivery Time | 3 days | 7 days | 10 days |
Number of Revisions | 1 | 1 | 2 |
AI Model Integration | |||
Batch Normalization | - | - | - |
Database Integration | - | - | - |
Detailed Code Comments | - | - | - |
Image Upscaling | - | - | - |
MLOps | - | - | - |
Model Deployment | - | - | - |
Model Documentation | - | ||
Model Monitoring | - | - | |
Model Testing & Optimization | |||
Model Tuning | - | - | |
Natural Language Processing | - | - | - |
NLP Tokenization | - | - | - |
Pre-Training | - | - | - |
Prompt Engineering | - | ||
Setup File | |||
Source Code |
Frequently asked questions
About Fernando
AI Marketing Automation | Klaviyo, GHL, HubSpot | n8n & AI Agents
Atibaia, Brazil - 12:44 am local time
40+ clients over 4 years, across ecommerce and B2B.
I am a marketing operator first. AI and automation are how I get the result, not what I sell you.
Most freelancers in this niche sell you a stack: flows that send, funnels that route, and nobody able to say which part made money. I build the revenue system and the reporting that proves what it earned.
WHERE THE REVENUE USUALLY IS
→ Klaviyo: welcome, cart, post-purchase and winback flows, plus segmentation that stops you mailing people who never buy again
→ GoHighLevel: funnels, pipelines and speed-to-lead follow-up, so paid traffic stops leaking between the ad and the booked call
→ HubSpot: lifecycle stages, lead scoring and workflows that make the pipeline report trustworthy
→ AI agents in n8n and Make: list hygiene, segment research, copy drafts, reply triage and weekly reporting, running daily without a hire
→ Deliverability: authentication, warm-up, list hygiene and sender reputation, fixed before I add volume, never sold as an upsell
RECENT WORK
→ High-ticket travel consultancy with near-zero online sales: I built the new-lead and reactivation automations. That channel now drives 6-figure ARR.
→ Silicon Valley tech company (high-8-figure ARR): I built the full HubSpot lead-tracking system that gave the founder his first clear pipeline view. He scaled sales and outreach on it, and that initiative now drives 7-figure ARR.
→ TikTok Shop, return center and Amazon MCF: I built the end-to-end operations automation that replaced manual work and saves $30k/month in payroll.
IF SOMEONE ELSE BUILT IT
I take over half-finished and broken Klaviyo, GoHighLevel, HubSpot and n8n builds. Every rescue starts with a written audit before I touch anything, then fixes run in order of what is leaking money.
HOW I WORK
→ Fixed price on fixed scope, in writing. Scope changes get re-quoted before I build, never after.
→ I tell you the monthly API and tool cost before I build, then design to stay under it.
→ Every build ships with retry logic, failure alerts, a Loom walkthrough and a one-page runbook. You should never learn from a customer that an automation died.
→ It all lives in your accounts. You can end the engagement and nothing breaks.
→ After launch: take it in-house with the documentation, or keep me monthly to monitor and extend.
Fluent in English and Portuguese. I serve US/EU teams and Brazilian brands.
WHO THIS IS FOR
Ecommerce and B2B teams already making money who want more from the traffic and the list they have. Not a fit if you want the cheapest bid, or one Zap you could build yourself in an afternoon.
Message me with (1) the platform you are on, (2) the number you want to move, (3) your timeline. Within 24 hours you get back a written read on what I think is actually broken and a fixed price. This is free whether or not we work together.
Steps for completing your project
After purchasing the project, send requirements so Fernando can start the project.
Delivery time starts when Fernando receives requirements from you.
Fernando works on your project following the steps below.
Revisions may occur after the delivery date.
Map the workflow and guardrails
Define triggers, tools, decisions, approvals, failure handling and success criteria.
Build and integrate the agent
Configure prompts, n8n or Make logic, APIs, retries, logging and human gates.