This study develops a reproducible Python-based data-analysis pipeline for processing open collider data and performing likelihood-based invariant-mass fitting. Z-boson reconstruction using ATLAS Open Data is used as a scientific-computing case study. The workflow integrates Jupyter Notebooks, Uproot, NumPy, pandas, Matplotlib, and the zfit fitting framework in a modular sequence comprising ROOT-file ingestion, event-variable access, lepton-quality filtering, vectorized event selection, invariant-mass construction, visualization, and statistical model fitting. Events containing electron or muon pairs were filtered using trigger, isolation, impact-parameter, pseudorapidity, transverse-momentum, charge, and same-flavor requirements. The selected dilepton mass distributions were fitted with a Double Crystal Ball model using zfit. The fitted mass parameter was 90.459 GeV for the electron channel, 90.823 GeV for the muon channel, and 90.678 GeV for the combined dilepton sample. The principal contribution is the organization of high-energy-physics data processing as a transparent and reusable scientific Python workflow rather than a new particle-physics measurement. The case study demonstrates how open ROOT-formatted datasets can be processed and statistically analyzed without building the complete analysis around a monolithic ROOT/C++ environment. The proposed pipeline supports notebook-level reproducibility, modular replacement of processing stages, and adaptation to other open scientific datasets with event-based structures.