Andres Ovidio Restrepo Rodriguez

Solve NLP and Speech challenges using state-of-the-art machine learning techniques. • Research and implement Deep Learning models for Text-To-Speech (TTS), generating synthetic voices with Colombian, Ecuadorian and Mexican accents, using Python, TensorFlow, PyTorch, and ONNX. GIT, DVC. • Research and implement Deep Learning models for Speaker Recognition, using Python, TensorFlow, PyTorch, and ONNX. • Deploy models in cloud environments (GCP and AWS) through APIs (Flask and FastAPI), MongoDB and Docker.

Part-time Researcher Professor, in fields such as Machine Learning, Data Mining and Natural Language Processing. • Principal researcher of the Research Project “Learning analytics model based on emotional metrics for second language learning mobile applications”. • Professor of the subjects of language processing, machine learning and data mining. • Collaborating researcher in the research project "Natural language processing for the evaluation of semi-automatic tasks", in collaboration with UNIR Spain.

Migration, implementation, and optimization of data pipelines and ETL, supported by technologies such as: Scala, Spark, PySpark, Maven, Jenkins, Artifactory, Git and Hadoop. • Design, specification and generation of meshes for scheduling the execution of data processing jobs using Control-M. • Support for clients in solving problems and bugs in Machine Learning projects.

• Develop the comparison of own and pre-trained convolutional neural network models through transfer learning for the classification of satellite images. • Implement algorithms for satellite image fusion under a heterogeneous CPU/GPU architecture. • Develop a software for creating, definition and training convolution neural networks for image classification. • Development a software for application of image segmentation techniques and their evaluation using similarity indices.

Master's thesis with application for laureate thesis.

Laureate Degree Project