People / Students / Class of 2027
Zhuoru (Nykole) Liu earned her bachelor’s degrees in Computer Science and Physics from the University of California, Berkeley. Her background combines software engineering, machine learning, and data systems with hands-on experience building pipelines for complex, noisy, real-world datasets. Through coursework in machine learning, natural language processing, computer architecture, and internet architecture, she developed strong technical skills in Python, SQL, C, Java, pandas, scikit-learn, PyTorch, TensorFlow, and data visualization.
At Berkeley’s Space Sciences Laboratory, Nykole developed an automated radio burst detection program for spacecraft radio spectrum data. The project addressed a major scalability challenge, which is that manually inspecting time-frequency spectrograms was too slow and inconsistent for large volumes of mission data. To adapt a ground-based computer vision method for spacecraft observations, she redesigned the pipeline with log-frequency interpolation, quantile-based thresholding, duration filtering, and Hough transform extraction. Her system processed noisy spectrogram inputs, identified structured burst candidates, and generated a catalog of 668 potential solar radio burst events. She also contributed to NASA Habitable Worlds Observatory research by developing a Python data integration pipeline for 14,315 radial velocity records from 17 instruments across more than 33 years of observations, benchmarking detection results against 69 reported planets and completing 110 signal verification checks. This work contributed to a paper accepted for publication in The Astronomical Journal.
Nykole has also applied her data science skills in industry-facing projects. As a Large Language Model and Data Intern at Global Key Advisors, she cleaned and merged mentor datasets using pandas and enriched records with email verification, seniority, department, phone, and LinkedIn data. She also built a pipeline to transform seven years of startup evaluation scores for more than 900 companies into a standardized analytical dataset, extracting valid scoring ranges from 283 rubric questions with regular expressions and normalizing evaluator responses to comparable 0-1 scores. Across her work, Nykole is especially interested in data engineering, machine learning infrastructure, and applied AI systems that make messy, real-world data more usable. At Northwestern, she hopes to deepen her technical foundation and prepare for a career building scalable, interpretable, and industry-ready data science solutions.
