Alex Muñoz

Key Skills

Software

AWS SageMaker

Google Cloud Vertex AI

Labelbox

Label Studio

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Text

Top Task Types

Bounding Box

Classification

Computer Programming Coding

Freelancer Overview

I worked as a Computer Vision Engineer, creating and labeling datasets for drones (UAVs). My main task was drawing bounding boxes around fires, vehicles, and people in aerial images to train AI models. I also helped improve these models using Python and techniques like transfer learning. Beyond labeling, I worked on making data processing faster with Docker and Google Cloud. My role combined attention to detail (for accurate labels) with technical skills (like optimizing pipelines) to build better AI systems.

Labeling Experience

Computer Vision Engineer

Google Cloud Vertex AIImageBounding Box

As part of this initiative, I contributed to two distinct AI models deployed on UAVs (drones) for patrolling Chilean forests. The first model detects wildfires in real-time, triggering instant alerts to an online platform. The second model identifies motorized vehicles and humans in the vicinity, aiding in both rescue operations and culprit identification. I actively participated in dataset creation (labeling bounding boxes for fires, vehicles, and people) and later trained both models using data augmentation and transfer learning techniques. The system’s end-to-end functionality ensures rapid wildfire response while simultaneously gathering critical visual evidence, enhancing both environmental protection and law enforcement efforts.

As part of this initiative, I contributed to two distinct AI models deployed on UAVs (drones) for patrolling Chilean forests. The first model detects wildfires in real-time, triggering instant alerts to an online platform. The second model identifies motorized vehicles and humans in the vicinity, aiding in both rescue operations and culprit identification. I actively participated in dataset creation (labeling bounding boxes for fires, vehicles, and people) and later trained both models using data augmentation and transfer learning techniques. The system’s end-to-end functionality ensures rapid wildfire response while simultaneously gathering critical visual evidence, enhancing both environmental protection and law enforcement efforts.

2023 - 2024

Education

Work History