In Nequi, I collaborate within a team to support various processes concerning risk analytics. My role involves upgrading them and continually striving for best practices and ongoing development, particularly in AWS cloud environments. Additionally, I contribute to Airflow orchestration and data management to ensure seamless operations and optimal utilization of resources.

During my time at Nequi, I had a pivotal role in implementing critical infrastructure and technological solutions that drove efficiency and innovation in the AWS cloud environment. My responsibilities included:- Contributing to the management and deployment of machine learning models, stored in ECR and executed through ECS and Fargate, ensuring the availability and performance of AI-driven applications.- Adapting and optimizing Docker containers for linear regression models, enhancing application portability and performance.- Actively participating in parameterization and orchestration for model execution using Airflow, ensuring effective management of scheduled tasks and complex processes.- Developing ETLs (Extract, Transform, Load) using Spark in AWS Glue, enabling the transformation of raw data into valuable insights.- Collaborating on the implementation of the necessary infrastructure to enable a serverless AWS Lambda service, which allowed for highly scalable and cost-effective function execution.

As a member of the Deodata team, I worked with a variety of technologies, including: Docker, YOLO, Amazon Web Services (AWS), Python, Pandas, SQL databases and Qt graphic interfacesMy role as a machine learning engineer involved:Enhancing an artificial intelligence algorithm hosted in an AWS-based API.Developing tools to create a more user-friendly interface for Deodata platform users.Extracting and transforming data from videos to train our convolutional neural networks.Deodata is designed to collect data for mobility studies from videos using artificial intelligence in a safe, fast, and efficient manner.

- Develop and run Python scripts to analyze data on vehicle counts to improve urban mobility- Utilize AWS and local resources to gather and process data- Test data for viability and assess its usefulness in optimizing neural network performance- Collaborate with cross-functional teams to develop and implement data-driven solutions

In the communications lab, I maintained and inventoried the equipment and supplies.I was in charge of a course about teaching robotics and programming utilizing the LEGO Mindstorms kits, as well as inventory and maintenance on them. We worked on a variety of projects, including segways, edge-followers, color classifiers, and sumo cars.