Backend Developer at MYEM (My Energy Manager)Responsibilities:- Collaborated with the team to design effective architectures and deployment strategies.- Developed microservices using Python, FastAPI, and Nameko framework.- Utilized Docker containers and Kubernetes for deployment and management.- Conducted unit and integration tests using Pytest for quality assurance.

As a part of a project titled "Prediction of drug-target interactions using deep learning models", I contributed to the development of a classification system aimed at identifying the interactions between drugs and its targets (proteins). My specific role in the project involved generating characteristics for drugs and targets, augmenting imbalanced datasets by generating fake drug-target data with similar characteristics and properties to the real-world data. To achieve that, I utilized various machine learning and deep learning architectures, including variational auto-encoders.The overall objectives were : - Generate characteristics for drugs and targets by using Collaborative Matrix Factorization (CMF).- Design and development of a deep generative model (VAE) that generates more training data for drugs and targets to address the problem of data imbalance between known-interactions and unknown-interactions data.- Propose an Artificial Neural Network for the classification process of interactions data. - Evaluate the quality of the generated data and its effect on the classification results.