# Gabriel Lopez Romero — Full Curriculum & Technical Portfolio > Backend Developer and Computer Science Student (UNCUYO) focused on backend development, cloud infrastructure, and applied machine learning research. ## Contact & Links - Website: https://lopezromerogabriel.vercel.app/ - Email: gabriellopezromero@gmail.com - GitHub: https://github.com/gabichulas - LinkedIn: https://www.linkedin.com/in/glopezromero/ - ORCID: https://orcid.org/0009-0002-9168-5350 - Google Scholar: https://scholar.google.com/citations?user=Py7s7CoAAAAJ - Resume PDF: https://github.com/gabichulas/portfolio/raw/main/Gabriel_Lopez_Romero_Resume.pdf --- ## About Me & Philosophy Computer Science student at Universidad Nacional de Cuyo (UNCUYO, Mendoza, Argentina) and backend developer. I like understanding how software works under the hood: network communication, OS-level interactions, and efficient resource usage. I primarily work with Python (FastAPI), I'm actively learning Go, and I'm very comfortable in Linux and Docker environments. Alongside software development, I do academic research at my university, where I develop Machine Learning and evolutionary algorithm approaches for cloud resource optimization and autoscaling. --- ## Technical Skills Breakdown ### 1. Backend Development & Languages - Python: High-performance backend development, asynchronous I/O, binary stream processing, data pipelines, FastAPI. - Go: Concurrency primitives (goroutines, channels), CLI tools, network services, microservices. - FastAPI: Asynchronous REST API development, Pydantic validation, automatic OpenAPI/Swagger documentation. - Java: Object-oriented architecture, scientific computing with MOEA Framework for multi-objective optimization algorithms. - Bash & Scripting: Automation, process orchestration, Linux server administration, cron jobs, system pipelines. - Linux: File systems, processes, memory models, network sockets, systemd, permissions, IPC. ### 2. Databases & Messaging - PostgreSQL: Relational database modeling, indexing strategies, query optimization, transactions, ACID compliance. - MongoDB: Document-oriented NoSQL database modeling, aggregation pipelines, schema validation, high-throughput storage. - Redis: In-memory caching, key-value storage, pub/sub messaging, session storage, rate limiting. - AWS SQS: Asynchronous queue-based message decoupling, dead-letter queues, event-driven processing. ### 3. Cloud Infrastructure & DevOps - Docker: Multi-stage container builds, image size optimization, layer caching, container networking, microservices isolation. - Kubernetes: Pods, Deployments, Services, ConfigMaps, Ingress, cluster scaling, declarative state management. - Terraform: Infrastructure as Code (IaC), state management, modular infrastructure provisioning on cloud providers. - Google Cloud Platform (GCP): Cloud Functions, Cloud Run, Cloud Storage, Pub/Sub, IAM & Workload Identity Federation (OIDC). - Amazon Web Services (AWS): EC2, S3, SQS (distributed message queuing), IAM, Lambda. - GitHub Actions: Automated CI/CD pipelines, linting, testing, Docker image building, secure OIDC deployment authentication. - Ansible: Configuration management, automated remote server provisioning, idempotent playbooks. - Serverless Architectures: Event-driven cloud compute, stateless execution, auto-scaling triggers. ### 4. Tools & Observability - Git: Version control, branching strategies, rebasing, bisecting, code reviews. - Prometheus: Metrics collection, time-series data scraping, alerting rules, exporter configurations. - Grafana: Dashboard creation, real-time visualization of systems telemetry, logs integration. - Loki: Log aggregation, LogQL query language, centralized log analysis. --- ## Featured Software Projects ### 1. pysync — Pure Python rsync Implementation - Description: Implementation of the rsync algorithm built entirely in pure Python from scratch. Handles binary data streaming, differential rolling checksums, MD5 block hashing, SSH secure transport, and buffer manipulation for efficient remote and local file synchronization. - Technologies: Python, SSH, Linux, Bash, Networking, Hashing, Binary Buffers & Streams. - Repository: https://github.com/gabichulas/pysync ### 2. f1-dataops — Serverless Formula 1 Telemetry Pipeline - Description: A 100% serverless, event-driven data pipeline deployed on Google Cloud Platform to ingest, process, and store Formula 1 live telemetry and historical race data. Features an end-to-end CI/CD deployment pipeline with GitHub Actions, Terraform Infrastructure as Code, Docker containers, and passwordless OIDC & IAM authentication. - Technologies: Terraform, GCP, GitHub Actions (CI/CD), Serverless, Python, Docker, OIDC & IAM. - Repository: https://github.com/gabichulas/f1-dataops ### 3. Demix — Deep Learning Audio Source Separation - Description: Comparative study of U-Net vs. Vision Transformer (ViT) deep learning architectures for audio source separation. Packaged into a containerized FastAPI backend inference microservice with Docker for real-time audio stem demixing. - Technologies: Deep Learning, Audio Processing, Research, FastAPI, Docker, PyTorch. - Repository: https://github.com/gabichulas/demix ### 4. PicVault — Facial Recognition Image Management Platform - Description: Intelligent media management and delivery platform for event photographers. Attendees can instantly find, filter, and claim their high-resolution photos from massive event galleries using deep facial recognition and feature embedding search. - Technologies: Computer Vision, Deep Learning, Image Processing, Django & DRF, React, Docker, PostgreSQL. - Repository: https://github.com/gabichulas/picvault --- ## Academic Research & Scientific Publications ### 1. Neuroevolution for Generating High-Quality Autoscaling Policies on Scientific Workflows - Authors: Gabriel Lopez Romero et al. - Status: Poster Submission for CARLA 2026 (Latin American High Performance Computing Conference) / Evolving to Full Paper. - Keywords: Cloud Autoscaling, Neuroevolution, Deep Learning, Reinforcement Learning, Distributed Systems, FinOps, Multi-Objective Optimization. - Full Abstract: "Scientific workflows are complex sequences of computational tasks designed to manage large-scale experiments and simulations. These workflows are typically executed in Cloud Computing environments, where achieving optimal performance depends on effective autoscaling strategies. Two prominent approaches for workflow autoscaling are Reinforcement Learning (RL) and Evolutionary Algorithms (EAs). Previous studies have combined these techniques by using EAs to optimize the initial policies and Q-tables of RL agents, enabling the agents to start from a pre-optimized behavior rather than learning entirely from scratch. Building on this idea, this work proposes a novel framework based on Neuroevolution. In this approach, neural networks are treated as individuals within a population and evolved through iterative cycles of fitness-driven variation in the parameter space. The evolved neural networks are then used to generate the initial Q-tables for the RL agents, providing a more effective initialization. Experimental results demonstrate that the proposed approach outperforms previous methods in two significant ways. First, considering both execution time (makespan) and monetary cost, the Neuroevolution-based approach produces Pareto fronts that dominate those obtained using conventional Evolutionary Algorithms. This improvement is reflected in hypervolume increases ranging from 7 times to more than 700 times across four benchmark scientific workflows. Second, the minimum makespan and cost values achieved through traditional Reinforcement Learning are located close to the Pareto fronts generated by the Neuroevolution approach, indicating that the proposed method effectively identifies high-quality trade-off solutions before the RL optimization process begins." ### 2. Efficiency and Plasticity Metrics for Database Technology Evaluation - Authors: Gabriel Lopez Romero et al. - Status: Technical Report & Paper in Conference Proceedings of Jornadas de Jóvenes Investigadores (AUGM JJI). - Keywords: Database Systems, Graph Databases, Benchmarking, Performance Metrics, Data Engineering. - Full Abstract: "Research focusing on formulating and evaluating quantitative efficiency and plasticity metrics for formal benchmarking and performance evaluation of database technologies across relational, document, and graph database engines." - Proceedings Link: https://grupomontevideo.org/site/wp-content/uploads/2026/02/Num3EX.pdf - Google Scholar: https://scholar.google.com/citations?user=Py7s7CoAAAAJ --- ## Professional Experience ### Backend Developer (Freelance) - Period: November 2025 - Present - Description: Developed backend services, automation tasks, and data processing pipelines using Python (FastAPI and Django), relational (PostgreSQL) and non-relational (Redis) databases, and Docker containerization. Prioritized code modularity, endpoint throughput, and architecture robustness. - Core Skills: Python, FastAPI, Django, Docker, PostgreSQL. ### Research Assistant (LABSIN UNCUYO) - Period: June 2025 - Present - Description: Research and design of AI-driven cloud resource management and autoscaling policies for massive scientific workflows. Work focuses on the intersection of artificial intelligence, multi-objective evolutionary optimization (MOEA Framework in Java), neural networks, and cloud computing infrastructure. - Core Skills: Java, MOEA Framework, Neural Networks, Multi-objective Optimization, Scientific Research. ### IT Support & Infrastructure Diagnostics (Freelance) - Period: December 2020 - Present - Description: Hardware and software diagnostics, Linux and Windows OS integrity, backup disaster recovery planning, and systems troubleshooting. - Core Skills: Hardware & Infrastructure Diagnostics, Incident Resolution & RCA, Disaster Recovery & Data Integrity. --- ## Education - Degree: Licenciatura en Ciencias de la Computación (B.Sc. in Computer Science) - Institution: Universidad Nacional de Cuyo (UNCUYO), Mendoza, Argentina - Period: 2022 - Present (In Progress) --- ## Available Machine-Readable Endpoints - OpenAPI Spec: https://lopezromerogabriel.vercel.app/openapi.json - Profile API: https://lopezromerogabriel.vercel.app/api/profile - Projects API: https://lopezromerogabriel.vercel.app/api/projects - Research API: https://lopezromerogabriel.vercel.app/api/research - Skills API: https://lopezromerogabriel.vercel.app/api/skills - LLM Index: https://lopezromerogabriel.vercel.app/llms.txt - Full LLM Context: https://lopezromerogabriel.vercel.app/llms-full.txt - Sitemap: https://lopezromerogabriel.vercel.app/sitemap.xml