Pesquisador e doutorando em Ciência da Computação | AI Systems · Machine Learning · Data · Architecture · Automation
Artificial Intelligence • Machine Learning • Data Engineering • Software Architecture • Distributed Systems • Cloud • MLOps • Intelligent Automation
Sou Mestre em Engenharia Elétrica e doutorando em Ciência da Computação pela UFSCar, com atuação em pesquisa aplicada e desenvolvimento de sistemas na interseção entre Inteligência Artificial, Machine Learning, Engenharia de Dados, Arquitetura de Software e Automação Inteligente.
Minha principal profundidade técnica está em AI/ML e Data. A partir desse núcleo, meus projetos avançam por backend engineering, sistemas distribuídos, cloud, DevOps, MLOps, IoT, robótica, controle e automação.
O trabalho vai de modelos de Machine Learning, RAG e sistemas multiagente a APIs, microsserviços, arquiteturas orientadas a eventos, plataformas de dados, infraestrutura cloud-native, observabilidade e governança de IA.
Dados → Inteligência → Decisão → Execução
Meu objetivo é ir além do modelo isolado e trabalhar o sistema completo em que ele opera: arquitetura, integração, escala, observabilidade, segurança, governança e operação.
🔬 Pesquisa aplicada em Inteligência Artificial, Machine Learning, sistemas inteligentes, automação e aplicações orientadas a dados.
🛠️ Desenvolvimento de projetos autorais em AI/ML, Data Engineering, Agentic AI, RAG, SaaS, arquitetura de software, cloud e automação.
🚀 Evolução de projetos de portfólio para arquiteturas mais completas, com System Design, DDD, microsserviços, mensageria, Kubernetes, MLOps, observabilidade e AI Security.
🎯 Foco profissional em AI/ML Engineering, Data Engineering, AI Engineering, Machine Learning, Software/Solution Architecture e Intelligent Automation.
Minha profundidade principal está em AI/ML e Data Engineering. A amplitude vem de arquitetura de software, sistemas distribuídos, cloud/platform engineering e sistemas inteligentes.
| Software & Solution Architecture | AI / ML / Data — Profundidade | Cloud, Platform & Systems |
|---|---|---|
| DDD • SOLID • Design Patterns | Machine Learning • Deep Learning | AWS • Docker • Kubernetes |
| Clean / Hexagonal Architecture | Reinforcement Learning • GenAI | Terraform • IaC • CI/CD |
| Microservices • Event-Driven | RAG • Agentic AI • MLOps | GitOps • Service Mesh |
| Kafka • RabbitMQ • APIs | Data Engineering • Model Evaluation | Observability • Reliability |
| System Design • ADRs • RFCs | AI Security • Adversarial ML | IoT • Robotics • Automation |
flowchart TB
subgraph B["AMPLITUDE DE ENGENHARIA"]
direction LR
A["Software & Solution Architecture<br/>DDD • SOLID • Design Patterns<br/>Microservices • Event-Driven<br/>System Design • APIs"]
C["Cloud & Platform Engineering<br/>AWS • Docker • Kubernetes<br/>Terraform • CI/CD • GitOps<br/>Service Mesh • Observability"]
D["Intelligent Systems<br/>IoT • Robotics • Control<br/>Automation • Embedded Systems<br/>Signal Processing"]
end
B --> E["AI / ML / DATA"]
E --> F["Machine Learning • Deep Learning • Reinforcement Learning"]
F --> G["Generative AI • RAG • Agentic AI • Multi-Agent Systems"]
G --> H["Data Engineering • MLOps • Model Evaluation"]
H --> I["AI Security • Adversarial Machine Learning"]
Esse perfil T-shaped me permite aprofundar modelos, pipelines e sistemas de IA sem perder de vista o ambiente completo no qual eles operam.
Plataforma full-stack de Dados & IA para Customer Intelligence e principal projeto de arquitetura do portfólio.
Integra Data Engineering, Lakehouse, Data Warehouse, MDM/Golden Record, Machine Learning, RAG, sistemas multiagente, governança e infraestrutura cloud. A evolução arquitetural incorpora DDD, microsserviços, arquitetura orientada a eventos, Kafka/RabbitMQ, Java/Python, APIs distribuídas, Kubernetes, Service Mesh, Data Mesh, MLOps e AI Security.
Python • Databricks • Snowflake • Spark • dbt • MLflow • LangGraph • Agno • MCP • RAG • AWS • Kubernetes • Terraform
Plataforma de governança, privacidade e avaliação de sistemas de IA sob a LGPD, com PII detection, políticas declarativas, geração de RIPD, fairness audit, red-teaming, differential privacy e verificação formal.
A evolução adiciona uma camada de Adversarial Machine Learning para avaliação de ataques, comportamento sob perturbações, riscos de modelos e estratégias defensivas.
Python • FastAPI • Streamlit • Machine Learning • AI Governance • Privacy • Compliance • Red Teaming • Adversarial ML
Plataforma full-stack de inteligência de receita e retenção para e-commerce, integrando ETL, Data Warehouse, Machine Learning para churn/forecast/recomendação, API REST/GraphQL e dashboard Next.js.
A evolução arquitetural explora Event-Driven Architecture, processamento assíncrono, mensageria, serviços distribuídos e observabilidade.
Python • TypeScript • Next.js • Node.js • GraphQL • Machine Learning • Data Engineering • Distributed Systems
Plataforma de automação de infraestrutura AWS com abordagem Terraform-first, gerenciamento de configuração via Ansible/PowerShell, controller Python, CI/CD, segurança, observabilidade e mecanismos de self-healing.
O projeto reproduz o ciclo de uma plataforma enterprise: provisionar → configurar → monitorar → detectar → diagnosticar → remediar → validar → registrar.
AWS • Terraform • Ansible • PowerShell • Python • GitHub Actions • CI/CD • Observability • Self-Healing
Projeto orientado à segurança de aplicações baseadas em LLMs, explorando RAG security, roteamento multi-modelo, red teaming, avaliação adversarial e observabilidade.
LLM Security • LLMOps • RAG Security • Red Teaming • AI Governance • Observability
Pipeline de IA para monitoramento e quantificação de incerteza regulatória, integrando ingestão documental, processamento semântico, embeddings, classificação com LLMs, índices temporais e dashboard.
Python • NLP • Embeddings • LLMs • Streamlit • Regulatory Tech • Time-Series Intelligence
Laboratório de segurança que cobre o ciclo completo — do pentest clássico ao Adversarial ML e à detecção de fraude por IA generativa. Principal projeto de AI Security do portfólio.
Implementa, do zero em NumPy/scikit-learn, ataques e defesas de Machine Learning: data e clean-label poisoning, evasão FGSM/PGD, transferability, model extraction/inversion, membership inference, adversarial training, defensive distillation, watermarking, DP-SGD e MLSecOps (drift/monitoring, incident response). Na frente de fraude por IA: voz clonada (MFCC/anti-spoofing), deepfake de vídeo, forense de imagem (ELA), phishing (URL/LLM-as-judge) e cadeia de custódia. Governança com MITRE ATLAS, OWASP ML Top 10, NIST AI RMF, ISO/IEC 42001 e EU AI Act. Cobertura validada por 20/20 testes E2E.
Python • scikit-learn • NumPy • librosa • OpenCV • Adversarial ML • MLSecOps • Deepfake Detection • AI Governance
| Projeto | Área | Foco técnico |
|---|---|---|
| Self-Evolving RL-PID AGV | Robotics / RL | Reinforcement Learning • PID • Adaptive Control • AGV • Adversarial RL |
| OmniMind AI OS | Agentic AI | Planning • RAG • Memory • Multi-Agent Systems • AI Infrastructure |
| Churn Intelligence | ML / Data | Churn • RAG • FastAPI • MLflow • Streaming • Model Evaluation |
| VisionGuard | Computer Vision | YOLOv8 • OpenCV • ResNet • Event-Driven Vision • Adversarial CV |
| AI Network Optimizer | Telecom / ML | O-RAN • xApp • Machine Learning • gRPC • Kubernetes |
| AuditFlow | Automation / BI | RPA • ETL • ML • Explainable AI • BI |
| RiskCredit | Credit Risk / ML | Classification • Explainability • Risk Modeling |
| Credit Score Predictor | ML + Cloud | AWS • Lambda • SageMaker • API Gateway • DynamoDB |
| Product Recommendation NLP + Graph | Recommendation Systems | DistilBERT • Graphs • PyTorch • NLP |
| Voice Anti-Spoofing | AI Security | DSP • Deep Learning • Voice Biometrics • PAD • Audio Deepfake Detection |
| AuraSense | AI + Signal Processing | DSP • Machine Learning • Sensor Data • Intelligent Systems |
| QAOA Portfolio | Quantum Computing | Qiskit • QAOA • QUBO • Portfolio Optimization |
| Agente-RAG | RAG / Automation | n8n • RAG • Google Drive • Telegram • Semantic Search |
| AgenteMCP | Agentic AI | MCP • n8n • LLMs • Dynamic Tool Selection |
| Multi-Agent Customer Insights | Multi-Agent AI | Research • Analysis • Strategy • Orchestration |
| Agente Calendar + HubSpot | Business Automation | HubSpot • Google Calendar • Supabase • WhatsApp • n8n |
| Agente Conversacional | Conversational AI | WhatsApp • GPT • Supabase • Memory • n8n |
| n8n | Intelligent Automation | Agentic AI • SaaS • CRM • Marketing • Workflow Automation |
| Engenharia Elétrica | Engineering R&D | IoT • Embedded Systems • Power Quality • Signal Processing |
| Analisador de DHT | Power Quality / IoT | ESP32 • Harmonics • THD/DHT • DSP |
| TCC — Smart Home Embedded | Embedded / IoT | ARM TM4C1294 • Webserver • ThingSpeak • Automation |
| IoT ESP32-CAM | IoT / Computer Vision | ESP32-CAM • PIR • Streaming • Home Security |
| Chatbot3IA | Conversational AI | Flask • Local ML • AI APIs • Chatbots |
| Heart Rate via Webcam | Computer Vision / DSP | rPPG • OpenCV • Digital Filtering • Spectral Analysis |
| Transfer Android | Automation | Python • ADB • File Transfer • Automation |
| YTDL4EDU | Automation / Education | Python • yt-dlp • Media Processing |
| IA Agentica | Agentic AI Infrastructure | Agents • Skills • MCP Servers • Git Submodules |
| Projetos Pessoais | Portfolio Hub | AI • Data • Automation • Research |
| Profile README | Portfolio | GitHub profile and technical positioning |
🔗 Ver todos os repositórios públicos no GitHub
A lista pública completa e atualizada está disponível em:
LangChain • LangGraph • Agno • CrewAI • OpenAI API • Claude API • RAGAS • DeepEval • Langfuse • MCP • A2A • Tool Calling • Embeddings • Vector Search
Machine Learning • Deep Learning • Reinforcement Learning • NLP • Computer Vision • Feature Engineering • Model Evaluation • Explainable AI • Predictive Modeling • Anomaly Detection • Adversarial Machine Learning
Ofensiva & Defensiva: Pentest • Reconhecimento (Nmap) • Brute Force • Malware Simulado • Linux Hardening • Defense in Depth • Incident Response
AI / ML Security: Adversarial ML • Data & Clean-Label Poisoning • Evasion (FGSM/PGD) • Model Extraction/Inversion • Membership Inference • Adversarial Training • Defensive Distillation • Model Watermarking • Differential Privacy (DP-SGD) • MLSecOps • Drift Monitoring
Fraude por IA Generativa & Forense: Deepfake Detection (vídeo) • Voice Cloning / Anti-Spoofing (MFCC) • AI-Generated Text & Phishing Detection • Image Forensics (ELA) • Chain of Custody • Digital Forensics
Threat Modeling & Governança: MITRE ATLAS • OWASP ML Security Top 10 • NIST AI RMF • ISO/IEC 42001 • EU AI Act • LGPD
ETL / ELT • Data Pipelines • Data Modeling • Data Warehouse • Data Lake • Lakehouse • Data Mesh • Data Products • Medallion Architecture • Spark • Databricks • Snowflake • Delta Lake • Streaming • Data Quality • Lineage • BI
AWS • Azure • Terraform • Infrastructure as Code • Docker • Kubernetes • Helm • Service Mesh • CI/CD • GitOps • MLflow • Model Registry • Model Serving • Drift Detection • Observability
Python • Java • TypeScript • SQL • C/C++ • FastAPI • Spring Boot • REST • GraphQL • gRPC • Microservices • DDD • SOLID • Clean Architecture • Hexagonal Architecture • Design Patterns • Kafka • RabbitMQ • Event-Driven Architecture • System Design
n8n • RPA • Workflow Orchestration • CRM Integration • WhatsApp Automation • API Integration • Business Process Automation • Agentic Workflows • Event-Driven Automation
Artificial Intelligence • Machine Learning • Reinforcement Learning • Distributed AI • Multi-Agent Systems • Robotics • Adaptive Control • Digital Signal Processing • Computer Vision • IoT • Embedded Systems • Optimization • Mathematical Modeling • Statistical Modeling • Monte Carlo Simulation • Quantum Computing • QAOA / QUBO
🎓 Doutorado em Ciência da Computação — UFSCar — em andamento
Pesquisa aplicada em sistemas inteligentes, Inteligência Artificial, Machine Learning, automação, robótica e aplicações orientadas a dados.
🎓 Mestrado em Engenharia Elétrica — UFSCar
Pesquisa envolvendo modelagem matemática, sistemas de engenharia, processamento de sinais, qualidade de energia, IoT e sistemas embarcados.
📚 Formação complementar
Data Science • Data Analytics • Software Engineering • Internet of Things • Industry 4.0
Domain-Driven Design • SOLID • Clean Architecture • Hexagonal Architecture • Design Patterns • Microservices • Event-Driven Architecture • Kafka • RabbitMQ • REST • GraphQL • gRPC • Docker • Kubernetes • Service Mesh • Infrastructure as Code • GitOps
System Design • Functional / Non-Functional Requirements • Capacity Estimation • High Availability • Scalability • Consistency • Resilience • Caching • Security • Observability • Disaster Recovery • Architectural Trade-offs
ADRs • RFCs • C4 Diagrams • API Contracts • Domain Models • Runbooks • Architecture Documentation • Engineering Standards
Linha transversal de pesquisa e desenvolvimento aplicada a diferentes tipos de modelos e sistemas inteligentes.
Adversarial Examples • Evasion Attacks • Data Poisoning • Model Extraction • Model Inversion • Adversarial Training • Red Teaming • LLM Security • RAG Security • Model Evaluation
| Projeto | Aplicação |
|---|---|
| CyberLab | Adversarial ML end-to-end • Poisoning • FGSM/PGD • Model Extraction/Inversion • Membership Inference • DP-SGD • MLSecOps • Deepfake/Voice/Text Detection • MITRE ATLAS • OWASP ML Top 10 • NIST AI RMF • ISO/IEC 42001 |
| ThemisAI | Governance gates • Model security • Adversarial evaluation |
| VisionGuard | FGSM • PGD • Adversarial patches • Computer Vision |
| RiskCredit | Feature manipulation • Evasion • Tabular ML |
| Self-Evolving RL-PID AGV | Sensor perturbation • State attacks • Adversarial RL |
| AegisLLM | Prompt injection • Jailbreaks • RAG security • LLM red teaming |
flowchart LR
A["AI / ML"] --> B["Data Engineering"]
B --> C["Software Architecture"]
C --> D["Distributed Systems"]
D --> E["Cloud & Platform"]
E --> F["MLOps"]
F --> G["AI Security & Governance"]
G --> H["Production AI Systems"]
Estou consolidando os projetos em torno de uma pergunta de engenharia:
Como transformar um modelo de IA que funciona isoladamente em um sistema inteligente escalável, observável, resiliente, seguro, governado e pronto para produção?
IA sem arquitetura vira demo.
Dados sem contexto viram ruído.
Modelo sem avaliação vira hipótese.
Sistema distribuído sem observabilidade vira investigação paranormal.
Automação sem integração vira tarefa isolada.
Meu foco é construir sistemas nos quais IA não seja apenas uma interface de conversa ou uma feature isolada, mas parte real da arquitetura do produto.