Debaditya Roy

Background

About Debaditya Roy

I hold a PhD and 5+ years hands-on experience in data-driven machine learning, deep learning and AI. I also have a 4+ experience as an industry consultant where I have worked as software consultant in individual capacity and also managed teams for different consulting projects (at PwC). I am skilled in AI research and adept at designing custom AI algorithms tailored to address specific problems. I use popular ML frameworks such as pytorch, tensorflow, sklearn to build and deploy AI models. I extensively use MLOps and CI/CD tools such as MLFlow and Github Actions to manage these ML pipelines effectively. I am passionate about building end-to-end ML architectures starting from data pipeline to model deployment. Presently I work at Qamcom Research Technology as ML Specialist, where I am engaged with building an AI platform for image processing and remote sensing. I am also responsible for the AI domain globally within the organization. Prior to my PhD I have worked as machine learning engineer at Krick Technologic Gmbh and predict.io: Machine learning, Deep learning, Federated Learning, Differential Privacy, Time-series, Computer Vision, IoT Sensors, Radar based ML, Git, MLOps, Algorithm, Data analysis, Backend development, Fullstack architecture, APIs, Software Design Patterns, Team Management: Python, Sql, Julia, Java : Tensorflow, Pytorch, Scikit-Learn, Keras, Numpy, OpenCV, MLFlow. ℎ : Logistic Regression, SVM, Clustering, Anomaly Detection, and PCA, Linear, Ridge, Lasso Regressions, Random Forest, Boosting, Time-series/Sequence models (CRF, HMM): Convolutional Neural Networks, Recurrent Neural Networks, Encoder-Decoder Architectures, U-net, GANs: Django, Flask : Git, DVC /: Github Actions : S3, Lambda, Sagemaker, Athena