Backend engineer focused on building scalable APIs, solving domain problems through clean architecture, and exploring the intersection of systems design and algorithms. I design databases, implement authentication workflows, and integrate AI into production systems.
I specialize in backend services and API design, with practical experience building everything from REST APIs to complex domain logic engines. My work spans layered architectures, database-driven business logic, and real-world problem solving—from HR automation to academic enrollment workflows. I'm particularly interested in how architecture decisions compound over time, and I actively experiment with different patterns: async request handling, stored procedures, modular monorepos, and face recognition at scale.
When I'm not building backends, I'm exploring algorithms through practical implementations, learning how graph theory translates to clustering problems, and understanding why certain technologies solve specific problems better than others.
-
Languages: Python, TypeScript, Go, Java
-
Backend & APIs: FastAPI, Express, PostgreSQL, SQLModel
-
Architecture: Layered services, modular design, clean separation of concerns, MVC patterns
-
Authentication & Security: JWT, OAuth2, bcrypt, role-based access control
-
Data & NLP: SQLAlchemy, spaCy, PDF extraction, semantic matching, face embeddings
-
Tools & DevOps: Docker, Docker Compose, pytest, Pyright, strict type checking
SIGRH+ — HR management platform built with FastAPI and PostgreSQL. Implements shift-aware payroll calculation with overtime detection, face recognition attendance, NLP-driven CV matching against job skills, role-based access control, and comprehensive audit logging. Features include employee lifecycle management, leave workflows, and an Ollama-backed AI assistant. A showcase of multi-domain service architecture with strict layering: controllers → services → models.
SIU Guaraní Clone — Academic enrollment workflow engine built with PostgreSQL and Go. Centralizes complex business rules in PL/pgSQL stored procedures: credit prerequisites, quota assignment, grading workflows, and state transitions. Demonstrates database-first design with trigger-based notifications and comprehensive error logging. Shows how to move domain logic closer to the data.
Backend Lab — Deliberate engineering sandbox comparing four API architectures: async FastAPI, layered expense tracker with JWT, TypeScript-validated Express API, and session-based auth flow. Useful reference for understanding trade-offs between sync/async access, token vs. session-based auth, and testing strategies across Python and Node.js stacks.
Human Clustering — Desktop Java application that groups people by affinity using graph algorithms: constructs weighted graphs from interest profiles, applies Kruskal-style minimum spanning tree, and partitions by connected components. Clean separation between domain models, algorithms, and Swing UI. Demonstrates end-to-end algorithm implementation from concept to interactive tool.
Lights-Out Game — 2D exploration game built with Java and Swing. Implements real-time game loop, camera-relative rendering, tile-based collision detection, inventory-gated progression, and event-driven audio. Shows how to structure interactive systems without external engines: update cycles, input handling, and layered rendering.
Deep systems design: how distributed services coordinate, what tradeoffs emerge between different database architectures, and how to make architectural decisions that scale across teams. I'm interested in database-driven domain logic, API design patterns that work at scale, and the practical side of NLP integration in production systems. Also actively learning Go for building orchestration layers and infrastructure tooling.