Vasileios K.

Data Scientist and ML engineer with 5+ years of experience in academia and industry with numerous publications and pipelines currently in production. In recent years, I have fully transitioned to industry applications across various disciplines working from early-stage research and design to production level systems. My skills span both statistical modelling, machine learning algorithm development and generative “AI” modelling from ideation, experimentation, testing and efficient production deployment and system integration. My passion lies in effectively communicating business needs in a simplified manner and delivering appropriate solutions using state of the art methods, understanding where systems fail, why they do so, and how to improve upon using my expertise in data. An outline of my expertise and frameworks which I have extensively applied over the past 5+ years is, Engineering / Development: API development, deployment and integration, Databases (Postgres, SQL, Pinecone, Elastic, Mongo), Cloud Environments (Azure, AWS, Run pod, Vast, GCP), Orchestration (Docker, Kubernetes), Machine Learning Model management (Weights and Biases), Data Mining (Selenium), CI/CD Machine Learning / “AI”: Natural Language Processing, Large Language Models, NER, RAG, Text/Token Classification, Image Classification, OCR, Object Detection, Generative Models (VAEs / GANs), Latent Diffusion Models, Deep Learning (CNNs, RNNs, LSTM, Transformer Models), Reinforcement Learning (Environments, Agents), Voice Cloning, Text-to-Speech/Speech-to-Text models, Image-to-Text/ Text-to-Image models, Model Interpretability (LIME, SHAP), Model Training/Fine-Tuning, Machine Learning Frameworks (Tensorflow, Torch) Statistics / Data Science: Bayesian Statistics, Classical Statistics, Hypothesis Testing, Experimental Design, Causality, Causal Inference, Gaussian Processes, Distribution Testing, Anomaly Detection, Time-Series and Forecasting, Dimensionality redaction (PCA, t-SNE), clustering algorithms, Big Data integration, GPU based data systems (RAPIDS)