Building real systems today while preparing to work on the safety of the systems that will shape tomorrow.

I am a software engineer from Brazil focused on turning ambiguous, real-world problems into reliable software.

My experience began in the health insurance domain, where I designed and built operational CRM, workflow automation, and AI-assisted systems used to support real business processes. This work required more than writing code: I had to understand how people worked, translate fragmented requirements into system behavior, handle sensitive information under Brazil’s LGPD, and make technical decisions in environments where reliability mattered.

My current work involves technologies and practices such as:

• Python, FastAPI, Django, and SQLAlchemy
• PostgreSQL, Redis, and Docker
• REST APIs and third-party platform integrations
• Authentication, authorization, webhooks, and secure secret management
• GitHub-based development workflows, testing, deployment, and debugging
• AI-assisted engineering and agentic development workflows
• Designing systems around real operational constraints

I am especially interested in the boundary between software engineering, artificial intelligence, security, and human decision-making.

My long-term goal is to contribute to empirical AI safety and security research, particularly in areas such as:

• AI evaluations
• Adversarial testing and red teaming
• Model and agent security
• Reliable research infrastructure
• Interpretability and behavioral analysis
• Robustness of increasingly autonomous AI systems

I am not presenting myself as a finished researcher.

My strongest evidence today is not a list of academic publications. It is my ability to enter unfamiliar territory, decompose complex problems, learn quickly, debug relentlessly, and turn ideas into systems that function in the real world.

I am now expanding that foundation through deeper study of machine learning, statistics, empirical research, and AI safety.

This profile is also a letter to my future self.

A reminder that expertise is not something I need to possess before attempting difficult work. It is something I can build by confronting problems that are larger than my current abilities, producing evidence, learning from failure, and continuing to expand.

I do not need to solve today every problem that the person I will become may face.

That person will have experiences, knowledge, resources, and capabilities that I do not have yet.

For now, my responsibility is to keep building.

• Complex problems become manageable when they are made concrete.
• Good engineering begins with understanding reality, not defending an abstraction.
• AI should expand human capability without making safety an afterthought.
• Ambition and intellectual honesty must coexist.
• A problem that exceeds my current ability is not necessarily beyond my future ability.

I can grow until I am capable of facing the problems that matter.