Ai_robotics | JOBATUS

Experiencia

AI Writer

Towards Data Science

may 2022 - Actualmente

- Writing of theoretical articles about the most used Reinforcement Learning algorithms to date.

- Explanation and implementation in code of Machine Learning algorithms, especially
focused on Reinforcement Learning.

- Willingness to help people outside the world of Artificial Intelligence to understand complex concepts in an easy and intuitive way.

AI Software Developer

Sherpa.ai

jul 2021 - Actualmente

- Build the Federated Learning platform oﬀered by the company from the initial stages, developed using Python and based on TensorFlow and PyTorch frameworks.

- Integrate the most widely used algorithms to date for their correct operation in a Federated Learning context.

- Coordinate the diﬀerent components of the platform through diﬀerent AWS services, such as Lambda, ECS, ECR, CloudWatch, SQS, Cognito, S3 or DynamoDB.

- Work on a joint project with NIH (National Health Institute) for the detection of rare
diseases using Computer Vision and Machine Learning.

Formación

MSc in Computer Science with Artificial Intelligence

University of York

oct 2020 - dic 2022

- Specialized in Robotics and Reinforcement Learning.

- Python 3.

- Data Structures and Algorithms.

- Machine Learning.

Degree in Industrial Technology Engineering

University of the Basque Country - Faculty of Engineering Bilbao

sep 2015 - sep 2020

- Member of the Intelligent Control Research Group.

- Specialized in Electrical Engineering.

- Automation and Control. Electrical Technology and Electrical/Mechanical Engineering.

- Physics.

- Algebra, Calculus, Numerical Methods.

Idiomas

English - Full Professional Proficiency

Spanish - Native Proficiency

Basque - Full Professional Proficiency

French - Limited Working Proficiency

Otros datos

Personal Project - Robotic Manipulator RLOA

- Creation of a framework for training Manipulator Robots in the Obstacle Avoidance task, using Reinforcement Learning.

- Implementation and application of the NAF (Normalized Advantage Function) algorithm for the manipulator training, based on PyTorch.

- Generation of a simulated training environment for the Robotic Manipulator using Pybullet.

Personal Project - App for Sudoku Solving via Camera

- Android App for solving sudokus from a photo.

- The picture of the sudoku board is transformed and parsed using OpenCV.

- The digits present in the cells of the Sudoku board are transformed into NumPy integers using a Convolutional Neural Network (CNN) that classifies the images of the digits from 1 to 9.

- The sudoku is solved using a from-scratch implementation of the Backtracking Depth-First-Search algorithm.

Personal Project - Autonomous Navigation of TurtleBot3-Burger platform with ROS

- Construction of the Turtlebot3 Robot.

- Robot set-up and testing of its operability.

- Implementation of an Autonomous Navigation System for the Robot, such that it can move freely through a previously recognized and SLAM-mapped environment.

- Based on ROS (Robot Operating System).