Sound Event Detection & Audio Classification for in-company clients. Developed a baseline using ResNet50 and progressed to state-of-the-art (SOTA) Transformer models like HTS-AT and BEATs. Orchestrated deployment of AWS services including EC2, Sagemaker, ECR, ECS, Lambda, and S3. Implemented fine-tuning, built a custom trainer, and designed a DSP pipeline to achieve an end-to-end solution.

Addressed students' inquiries, providing clear responses to enhance their understanding of course material. Reviewed and assessed deliverables, ensuring alignment with course objectives. Maintained consistent follow-up with students, offering ongoing guidance and support. Motivated students fostering a proactive learning attitude and helped overcome difficulties.

• Products review classification: preprocess and vectorize data using BoW, TF-IDF and word embedding (NLP). • Image Classification for E-Commerce: finetuning/experimenting on CNN model in AWS and improving the results by cropping the images from the dataset using object detection (Computer Vision). • Flask ML API: architecture with Flask/Redis to get the class of the uploaded image using ResNet50 (CNN). • Home Credit Risk Analysis: Regressor, RandomForest, LightGBM with RandomSearch on Kaggle data (ML). • Salary Prediction Model: using Regression models to estimate salaries with data obtained from NBA-API (ML). • Main technologies: Python, Numpy, Pandas, Scikit-learn, TensorFlow, Keras, Docker, AWS.

• Tailor-made planning for each client's business premises and their needs, including advice on the acquisition of equipment, set-up/configuration of the audio system and eventually more specific acoustic treatment.

• Mentoring is customized to the requirements of each student and his or her academic objective or subject matter. • From mathematics, algebra or physics for high school and university entrance.