Jan 2024 – May 2024 · 4 mos · Amsterdam, North Holland, Netherlands · Hybrid

• I worked on a cross-functional environment with an architect, product managers, and team members. Key responsibilities included:
• Contributed to the maintenance of 7+ terraform repositories setup for AWS infrastructure, handling essential services like SQS, IAM, ECR, and more.
• Orchestrated Kubernetes deployments tailored to the needs of various Dutch Magazine product brands such as Ouders van Nu, Autoweek, De Ondernemer, BusinessWise, and VTWonen.
• Supported front-end and back-end teams by provisioning infrastructure and contributing to CI/CD pipelines workflow development.
• Created a pilot for a new distributed logging system using FluentBit and AWS OpenSearch. It was optimized for cost management by considering a set of scenarios with an AWS management account, and a set of AWS client accounts as the logging source (EKS clusters).
• It follows the guidelines and policies of the DPG's cloud IT governance, and It is expected to result in 30%+ cost savings when in production.
• Increased security access to OpenSearch dashboards by introducing SAML-based authentication, allowing using corporate identities (Google accounts + One login as identity provider sources) and group membership for RBAC.
• Provide input to the Architectural Decision Records documentation about the logging strategy for the product brands platform team's management.

Jan 2023 – Dec 2023 · 11 mos

Feb 2022 – Jan 2023 · 11 mos

Oct 2020 – Jan 2022 · 1 yr 3 mos · Nijmegen, Gelderland, Netherlands

• Deployed applications on Azure Kubernetes clusters by implementing a
• Terraform-focused approach via Terraform Cloud Workspaces and GitHub
• actions.
• Implementing a cluster level logging for a system based on
• microservices and deployed on AKS. (Fluentbit + Loki)

Feb 2020 – Aug 2020 · 6 mos · Utrecht, Netherlands

• Infrastructure as Code via ARM templates, using Gitlab and Azure DevOps for
• CI/CD. Preparing and deploying workloads like VNets, Cloud Firewalls, Load Balancers and
• virtual Machines.

Dec 2018 – Jan 2020 · 1 yr 1 mo · Utrecht Area, Netherlands

• Providing an automated and secure workflow allowing software delivery in
• testing, and production environments.
• Managing a multi-stage CI/CD pipeline in Azure DevOps, Maintaining IaaC workflow with Terraform
• Responsible for maintaining Azure Kubernetes Service environments.

Jul 2017 – Jun 2018 · 11 mos · Medellín

• Message transmssions architecture between SOFA (https://www.sofa-framework.org/) instances for medical training
• SOME CONTEXT:
• There was an existing custom application to perform some training activities for medical students.
• It was developed on top of SOFA and added to it as a custom plugin.
• This application provided liver, heart, and bladder 3d models and used a tracking phantom system to perform actions to these organs (cutting, carving, clip attaching, probbing and deformation computation when peers interact) by implementing distributed computing approaches.
• P2P: Every peer, compute and store the status of the actions and send to other peers these events. Every peer update this status on its side
• Client-Server: A centralized server node where the status of the simulation is computed and transmitted to every client to be visualized
• This solution was already implemented, but these approaches had a network overhead environment problem. So I propose the following way for messages transmission:
• Decouple the exchange of messages between SOFA entities using ZeroMQ. In that sense I use SQS systems to send the messages by defining a SUB-PUB model having a python process (Server) receiving messages from a client (PULL) and redirect them to the other clients (PUBLISHING). It helped to reduce the overhead and improve the solution throughput.
• An API REST endpoints to store training sessions and peers data.
• Django Rest Framework was used to implement a data model to allow store and distribution patients, peers, and training sessions data with external applications.
• The message transmission among SOFA nodes Instances was limited (in their first stage) to data instrument position and data deformation objects through the Hybrid PUSH-PULL model implementation to PUBLISHER - SUBSCRIBE ZMQ pattern

Nov 2015 – Dec 2016 · 1 yr 1 mo · Medellín

• Neurorehabilitation Information System
• I Built a CRUD system to manage the patient information inside of rehabilitation sessions performed in people who have suffered strokes and have compromised mobility of their upper limbs.
• I Built REST backend to expose users, rehabilitation sessions, patient evaluation monitoring endpoints services to allow the interoperability between the clinical information system and Unity platform with the purpose to use this patient information in the video games development to support the rehabilitation sessions.
• Requirements specification
• Setup a deployment architecture to implement the Rehabilitation Information system and their endpoint services.
• Analysis and tool selection and use of PaaS platforms (Heroku, Amazon S3, EC2) for the architecture development solution.
• Construction and continuous deployment of the solution.

Aug 2014 – Nov 2014 · 3 mos · Medellín

• Advisor in the research group in Applied Engineering - CII in the project "Design and Implementation of Integrated Hospital Information System, interoperable and with support for Teleradiology and Teledermatology for Healthcare Providers under the modality of Telemedicine in the Department of Antioquia" in its initial stage.
• Implement a proof of concept of a health care system that can be applied to the HMO of the Adventist University Corporation. This serves as a general support to the processes of the health entities of level 1, taking into account the possibility of interoperability with other healthcare systems, allowing a unified clinical history of patients in relation to the work of teledermatology and teleradiology.
• I Built an exploratory analysis and technical Review SoA about the following topics:
• Electronic Medical Records
• Interoperability between health applications or clinical systems information
• Health Level Seven standard
• Develop or Teleradiology and Teledermatology in the e-Medicine context
• DICOM - DIGITAL Imaging and Communication in Medicine standard
• Analysis and tools/frameworks selection to support the Clinical Information System oriented to patients.
• Implement a GNUHealth-based Clinical Information System.
• Deploy of an Infrastructure as a Service platform server to install the GNU Health framework
• Document the installation and deployment process.

Jan 2011 – Jul 2011 · 6 mos · Medellin

• WebTV platform proof concept to access to educational contents VoD and live streaming.
• Sysadmin LInux-based and Windows Server 2008 servers infrastructure to support the different requirements services of the ARTICA projects

Mar 2008 – Apr 2009 · 1 yr 1 mo

• I worked at project of Telemedicine supported by the Universities CES, EAFIT, Engineering School of Antioquia and the Mayo Clinic.
• The project consists in to create a Grid computing platform for exchange and storage medical images in DICOM format between organizations.
• Grid computing architecture proposal based on the DICOM images storage requirements and Globus Toolkit and OGSA-DAI (OGSA-DQP) Grid Open Source middleware's
• Development of a PHP-based grid portal frontend application to interoperate with CES, EIA, and EAFIT DICOM medical images data sources
• As a result of this project, the team work produces the following publication: mantisGRID: A grid platform for DICOM Medical Images Management in Colombia and Latin America
• https://link.springer.com/article/10.1007%2Fs10278-009-9265-x

Feb 2008 – Mar 2013 · 5 yrs 1 mo

• - Sysadmin Linux-based servers infrastructure to support the different requirements services of the Networks and Distributed Systems laboratory projects and the different subjects of Computer Networks specialization.