Location: Amsterdam, Netherlands (2x on-site)

Duration: 12 months, Full-Time

Starting Date: ASAP

Company: Tarucca BV

IMPORTANT: This is a hybrid position (Amsterdam) and requires a valid EU passport (or a Valid Dutch Visa)

Join Tarucca BV’s R&D team as a Machine Learning Engineer Intern and help pioneer physics-informed AI solutions for submarine cable localization. You will work in a mature, challenging environment, converting maritime sensor data into actionable insights using advanced machine learning and signal processing techniques. This internship offers hands-on experience with real-world sensor data, error-correction modeling, and robust MLOps frameworks.

🎯 Key Challenges

• Physics-Informed Model Development: Build and refine ensemble ML models that integrate analytical electromagnetics and algorithms such as Random Forests, SVMs, and gradient-boosted trees to improve cable depth estimation.
• Algorithm Implementation & Experimentation: Develop baseline algorithms (polynomial fits, least squares) and progress to advanced architectures. Apply domain adaptation to achieve reliability across cable types and environments (sand, clay, high-voltage cables).
• MLOps Pipeline Contribution: Design validation & retraining pipelines for automated model improvement, and data/version management. Contribute to a closed-loop system using new labeled data for iterative enhancements.
• Signal Processing & Data Engineering: Employ signal processing techniques (e.g., PCA, DTW, wavelet analysis) to denoise and structure raw sensor time-series, optimizing for high-quality ML training inputs.
• Uncertainty Quantification: Implement methods (e.g., Monte Carlo dropout, Bayesian techniques) to quantify predictive uncertainty.
• Hyperparameter Optimization: Use Bayesian optimization and active learning to tune model hyperparameters and refine sensor selection to improve SNR.
• Collaborative Research: Work closely with the R&D team, code review, document model architectures, and present technical findings daily.

📋 Requirements (Must Have)

• Education: Master's degree in a relevant field (Computer Science, Statistics, Physics, etc.) OR Bachelor's degree in a relevant field with 2+ years of relevant experience.
• Machine Learning: Strong grasp of supervised algorithms, e.g., Random Forest, SVMs, and boosting methods.