Jakub Gil Email & Phone Number

Warsaw, PL

Warsaw, Mazowieckie, Poland

• Assisted in AI strategy formulation, focusing on aligning tech with business and achieving product-market fit
• Deployed the "AI Stylist" for major fashion brands, optimizing workflow, streamlining processes and enhancing product quality
• Onboarded non-technical team members (Product Designers, Customer Success, Sales) into the "AI Stylist" project, breaking down tech concepts into simple terms

• Trained and fine-tuned advanced neural networks for classification, segmentation and detection tasks (VGG, ResNet, YOLO, Faster R-CNN, Detectron2)
• Applied convolutional neural networks (AlexNet, LeNet, custom CNNs) in representation learning, enhancing e-commerce recommender systems (~40% cross-sell increase)
• Developed data pipelines for computer vision: scraping (Selenium, BeautifulSoup), preprocessing (GoogleVision, OpenCV), labeling (VIA), storage (EFS, S3, ElasticSearch)
• Used unsupervised learning for dimensionality reduction (PCA) and style clustering (K-Means)
• Developed multimodal neural networks for image and tabular data
• Customized PyTorch Lightning architectures, datasets, loss functions, batch samplers

Warsaw, Mazowieckie, Poland

• ML Models Training — scikit-learn, XGBoost, CatBoost, LightGBM, GLM, Logistic regression
• Data Analysis — Pandas, NumPy, SciPy and Data Visualization — Matplotlib, Seaborn, Plotly
• Conducted hyperparameter tuning using Optuna and/or GridSearchCV
• Evaluated model performance with MLflow and TensorBoard — ROC, AUC, RMSE, Accuracy
• Generated artificial data for minority class oversampling using GANs in TensorFlow Keras
• Managed relational databases with Python (SQLAlchemy), Toad, and Dbeaver

• Utilized AGILE/SCRUM, JIRA, Confluence and Perforce for software development

Master'S Degree, Quantitative Methods In Economics And Information Systems

Engineer'S Degree, Mechanical Engineering