Sebastian Felipe Alzate Corredor

DevOps & Cloud Training – Internal

• Completed hands-on training focused on cloud infrastructure, containerization, CI/CD automation, monitoring, and Kubernetes deployments
• Deployed applications to Azure Kubernetes Service using Helmfile, including ConfigMaps, health probes, ingress configuration, and subpath exposure
• Built containerized applications with Docker and Docker Compose and designed GitLab CI/CD pipelines for automated build, test, and deployment workflows
• Gained practical experience with AWS and Azure services, including IAM, VPC, EC2, ECS, S3, RDS, and AKS
• Used Terraform, Terraspace, and reusable modules to provision and manage AWS infrastructure through Infrastructure as Code

Technologies: Kubernetes, Docker, Docker Compose, GitLab CI/CD, Terraform, Terraspace, AKS, AWS, Azure, Gatus, Helmfile, YAML

MLOps Training – Internal

• Built end-to-end ML workflows covering data preparation, model training, deployment, monitoring, and model lifecycle management
• Developed distributed pipelines with Ray and orchestrated workflows using Prefect
• Implemented Bayesian hyperparameter optimization, experiment tracking, and model versioning using Optuna and MLflow
• Fine-tuned deep learning models with PyTorch and PyTorch Lightning and developed inference services using FastAPI, React, and Streamlit
• Automated model evaluation and monitoring with Evidently, including data quality, stability, and drift reporting

Technologies: Python, PyTorch, PyTorch Lightning, MLflow, Prefect, Ray, Optuna, Evidently, FastAPI, React, Streamlit

Software Development – BP Anansi Project

• Developed and maintained a financial modeling web application for economic data analysis
• Implemented new features and enhancements across the frontend and backend using Django and React
• Provided real-time support diagnosing user-reported issues, identifying root causes, and efficiently resolving urgent problems to minimize user disruption
• Collaborated with cross-functional teams to deliver high-quality software on time

Technologies: Python, Django, React, Docker, HTML, CSS, TypeScript

Intelligent Chatbot for Industrial Process Control

• Designed and implemented a cloud-based system to control and monitor virtual industrial processes through chatbot commands
• Integrated Watson Assistant with Node-RED to interpret natural-language inputs and trigger actions
• Connected chatbot logic to PLCs via Studio5000 and Allen-Bradley hardware to manage I/O signals
• Simulated real-time industrial operations in Factory I/O controlled via Telegram chatbot

Technologies: Watson Assistant, Watson IoT, Node-RED, Factory I/O, Studio5000

MathWorks Minidrone Competition Colombia 2019

• Designed a vision-based line-following algorithm for a virtual Parrot Minidrone using Simulink
• Processed live drone camera feed to detect and track a red guideline
• Developed autonomous navigation logic based on real-time visual input
• Advanced to the final in-person round using real hardware after the initial virtual simulation phase

Technologies: MATLAB, Simulink