Applied ML / Stats PhD. Full-time code Houdini, tricking software into doing what I want. Take a look at my circus ring!
My PhD focused on teaching ships to navigate themselves: specifically, getting reinforcement learning agents to handle the messy reality of inland waterways, where currents are unpredictable, data is noisy, and running aground isn't an acceptable failure mode.
Along the way, I also spent nine months as the founding backend engineer at a mobility startup, building real-world systems from scratch.
My blog is where I follow whatever happens to boggle my mind: autonomous vessels, signal processing, Fourier transforms, ML shenanigans, and whatever else catches my attention.
Three-part blog series exploring the denoising autoencoder for speech enhancement. Achieved improvements over reference methods with 69% increase in PESQ and 9% increase in STOI. Covers theory, architecture , training, and deployment.
Interactive dashboard exploring the EuRepoC global cyber incidents database, visualizing state-level cyber conflicts across the globe. Built with Plotly Dash to create an intuitive interface for understanding patterns in cyber warfare, featuring dyadic incident visualization, country-level analytics, and timeline insights into the evolving threat landscape.
Shaped the technical foundation as founding engineer for a mobility startup from zero to seed. Developed an automated market-intelligence pipeline to monitor market trends, enabling robust cost forecasts and strategic insights. Followed the ride for 9 months from zero to seed.
A modularized DRL framework for autonomous vessel control on inland waterways, featuring separate agents for high-level path planning (considering traffic rules and dynamic obstacles) and low-level path following. Outperforms traditional APF and PID controllers by 65% in obstacle avoidance while reducing control effort.
An open-source Python framework for extracting clean ship trajectories from noisy AIS big data. Uses maneuverability-dependent α-quantile-based filtering to handle technical inaccuracies and compliance issues in raw AIS records. Robustly extracts long, uninterrupted trajectories for maritime domain awareness and algorithm testing.
A bootstrapped Deep Q-Network (DQN) controller for autonomous vessel navigation on restricted inland waterways. Handles challenging conditions like high flow velocities and shallow banks on the Rhine. Trained in diverse, realistic environments and validated against real-world data, demonstrating superior adaptability compared to vessel-specific PID controllers.
Martin Waltz, Niklas Paulig, Ostap Okhrin
Expert Systems with Applications, Volume 274, May 2025
Niklas Paulig, Ostap Okhrin
Ocean Engineering, Volume 312, Part 2, November 2024
Niklas Paulig, Ostap Okhrin
Ocean Engineering, Volume 298, April 2024
I'm always interested in discussing research collaborations, new opportunities, or interesting ML / CS problems.