Software Developer specialized in building scalable web applications using Laravel, .NET, Node.js, React, and Python — helping organizations transform business requirements into maintainable, high-performance software solutions.
My approach combines backend development, API design, database engineering, and software architecture to deliver products that are reliable and built to scale. I enjoy technical challenges that generate measurable impact for users and organizations.
My background in bilingual customer service adds strong communication, stakeholder management, and problem-solving skills that translate directly into collaborative engineering teams.
- 📍 Medellín, Colombia — remote-ready
- 🎓 Studying Data Science & AI Engineering at ESEIT
- 🌐 Spanish (Native) · English (C1 Advanced — National Geographic Certification)
Backend
Frontend
Data & Tooling
Nov 2025 – Present · Medellín, Colombia
- Developed full-stack features using Laravel, C# .NET, Node.js, React, and SQL
- Built REST APIs and integrated third-party services
- Optimized database queries and contributed to schema design
- Collaborated through Git-based workflows in Agile/Scrum environments
- Delivered scalable, maintainable solutions aligned with business requirements
Jan 2025 – Nov 2025 · Medellín, Colombia
- Managed operational records across information systems with a focus on data accuracy
- Investigated data discrepancies and coordinated resolutions across multiple stakeholders
- Handled high volumes of workflows and documentation in English and Spanish
- Built strong foundations in analytical thinking, process documentation, and information management
🎟️ NovaPass API
REST API for theatre ticketing and event management. Multi-role system (customer / admin / seller / scanner), JWT auth with token blacklisting, Google OAuth2, rate limiting, n8n workflow automation, and fully automated CI/CD pipeline to VPS via GitHub Actions → ghcr.io.
C# ASP.NET Core PostgreSQL MongoDB Docker JWT GitHub Actions
🛡️ Poison AI
Platform that detects poisoned, backdoored, or adversarially contaminated AI training datasets before they reach production. Fine-tunes small models (250M params) in isolated sandboxes and analyzes weights, loss curves, and label disparity to surface dangerous samples.
Python PyTorch HuggingFace Transformers PEFT NumPy scikit-learn Docker
| Data Science & AI Engineering | ESEIT — Jan 2026 – Present |
| Full Stack Developer Program | Scrimba — Jul 2025 – Present |
| C1 English Program | LCN — Oct 2021 – Dec 2024 |
| High School Diploma | Colegio Iberoamericano — Jul 2025 |
