Software Engineer · Backend & Full-Stack Developer · QA Automation Engineer

I'm a Software Engineer from Mexico who enjoys building full-stack applications, backend services, automation tools, and pretty much anything that lets me understand how software works behind the scenes.

Most of my work is around the JavaScript/TypeScript ecosystem, especially React, Next.js, Node.js, and NestJS. I also have experience building REST APIs with Java + Spring Boot and Python + FastAPI, as well as creating End-to-End automation with Playwright.

Recently, I've also been getting more into Linux, self-hosting, backend architecture, and infrastructure by building and managing my own Debian home server.

A self-hosted private cloud storage platform I'm building from scratch as an alternative to services like OneDrive and iCloud. The project is focused on understanding and implementing the architecture behind a cloud storage system, from file management and physical storage to server administration and automated backups.

Stack: NestJS · Next.js · TypeScript · Node.js · Prisma ORM · PostgreSQL · Debian 13 · Bash · Python

What I'm building:

• Building a REST API with NestJS for authentication, users, folders, files, uploads, downloads, and storage operations.
• Separating the logical filesystem stored in PostgreSQL from the physical filesystem stored on a dedicated HDD.
• Using UUID-based storage identifiers (storageName) for physical files while preserving user-facing filenames and folder structures as database metadata.
• Implementing hierarchical folder management using PostgreSQL relationships and recursive queries.
• Using Node.js Streams for file uploads and downloads to avoid loading entire files into memory.
• Supporting single-file and folder downloads, including ZIP generation for directory downloads.
• Building a Bash-based Management CLI accessed through SSH for monitoring system resources, storage, logs, services, backups, and maintenance tasks.
• Developing an independent Python/Bash backup automation system scheduled through cron, including backup creation, verification, and retention.
• Designing the system around limited hardware resources, making storage efficiency, memory usage, and server reliability important architectural considerations.

Current status: Authentication is complete, and the folder management system is nearly finished. File management, the frontend UI, responsive design, backup automation, and deployment are currently in development.

The long-term goal is to turn the project into a fully functional private cloud that my family and I can use to store and access files through a web interface, while maintaining complete control over the infrastructure and data. The platform is designed to eventually support both local-network access and secure public access.

A machine learning project I built to explore how historical traffic data can be used to predict accident severity.

Stack: Python · Scikit-Learn · Pandas · FastAPI · Java · Spring Boot · React · Leaflet

What I built:

• Processed and analyzed 7.7M+ US traffic accident records from 2016–2023 using Pandas and Scikit-Learn.
• Performed data cleaning, preprocessing, and feature engineering before training the model.
• Trained a Random Forest classifier with 83.61% accuracy for accident severity prediction.
• Built a FastAPI microservice responsible for loading the trained model and serving predictions through a REST API.
• Built a Spring Boot Backend-for-Frontend (BFF) to communicate with the prediction service and expose data to the frontend.
• Created an interactive React + Leaflet map for exploring accident information and geographic risk data.

This project was also my first real attempt at combining machine learning, microservices, Java, Python, and a web frontend into a single system.

A full-stack bug tracking and software quality management platform I originally built as my university residency project.

Stack: Next.js · TypeScript · PostgreSQL · Prisma · Playwright · Recharts

What I built:

• Implemented authentication and Role-Based Access Control (RBAC) for Administrator, Developer, and Tester workflows.
• Built ticket management with priorities, modules, statuses, assignments, and complete status history.
• Added role-specific permissions to control which users can create, modify, assign, close, or cancel tickets.
• Integrated Google Gemini to automatically estimate issue severity when it isn't manually provided.
• Built analytics dashboards with Recharts, including open-ticket metrics, reopened vs. closed issues, and Mean Time to Resolution (MTTR).
• Added PDF metric summary export

The goal was not only to build a CRUD application, but to model an actual software QA workflow with different responsibilities and permissions.

• GitHub Actions & CI/CD
• Linux system administration
• Backend and system architecture
• Self-hosting and infrastructure
• Building software that can actually run outside of my development machine :)

• Portfolio: fabianmadrid.dev
• LinkedIn: linkedin.com/in/sergio-acuña-59735336b
• Email: sergioac.madrid@hotmail.com