AI Automation Engineer
AI automation to scale operations without scaling headcount.
I am an AI Automation Engineer with 4+ years building LLM systems, agents, RAG and automations for companies that need to reduce rework, speed up decisions and turn manual processes into software.
Based in Porto Alegre, Brazil. Available remotely for Brazilian and international teams.
Services
I solve operational bottlenecks with applied AI.
Agents and multi-agent systems
Systems that execute complex tasks without constant supervision, using state, rules, tools and approval points.
- Stateful workflows
- Goal-oriented execution
- Human control for sensitive decisions
RAG and knowledge bases
Your team querying internal documents as if an analyst were available 24/7.
- Semantic search over PDFs, DOCX and internal data
- Responses grounded in retrieved context
- Architecture designed to reduce hallucination
API automations
Processes that depend on people today start running automatically with n8n, Make, FastAPI, webhooks and databases.
- Tool and system integrations
- Recurring routines automated
- Logs, alerts and traceability
Functional prototypes
Ideas validated in weeks, not months. First a useful version; then scale, security and hardening.
- AI and automation MVPs
- Fast product validation
- Technical foundation ready to evolve
Case studies
Projects that show technical ability and operational impact.
Main case
Iboti Advogados — Multi-agent legal assistant
Legal assistant with state memory, RAG and a knowledge base for analysis, research and document generation with human review.
Reduced legal research and first-draft time from hours to minutes
Lawyers can query precedents and draft documents without manual search
Stack: LangGraph, LangChain, Supabase, Python and OpenAI API
Prospecting
Instalight / Clinics — Automated lead generation
Prospecting tool with n8n, SerpAPI and an HTML app to find local clinics and generate qualified lists in a few clicks.
Removed repetitive manual research
Generated qualified lists from search criteria
Ready to evolve with CRM, enrichment and follow-up
Content
Daily content agent
Pipeline with n8n, SerpAPI, GPT-4o, Gmail and Supabase to monitor trends, deduplicate ideas and deliver briefings without manual curation.
Daily briefings delivered automatically
Deduplication to avoid repeated topics
Zero human curation in the operational loop
Stack
Technology used to deliver real systems, not just prototypes.
Python
LangGraph
LangChain
FastAPI
n8n
Make
Supabase
pgvector
OpenAI
Anthropic
OCR
Docker
REST APIs
How I work
I start with the bottleneck. Then I design the system.
I do not start by selling a chatbot. I start by understanding where there is rework, repeated decisions, manual documents, fragile spreadsheets or lack of traceability.
- Diagnosis: I map the current workflow, inputs, outputs and failure points.
- Functional prototype: I deliver a useful first version to validate value quickly.
- Integration: I connect APIs, databases, documents, automations and human approvals.
- Hardening: I improve quality, cost, logs, security and reliability.
When it makes sense to talk
I am useful when the problem already costs time, money or control.
Your team copies data, checks documents, fills systems or reviews information manually.
Documents, decisions and rules live across folders, spreadsheets, PDFs and conversations.
You want AI, but need architecture, validation, logs and human control to reduce risk.
Contact
Want to turn a manual process into a system?
Send me the process, the goal and where the rework happens today. I will give you a direct view of what can be automated and what should not be automated yet.