Worked on the JeffersonLab/coatjava repository to enhance the Kalman filter’s robustness by introducing dynamic magnetic field configuration within the AHDC engine. Replacing a previously hardcoded magnetic field value, the new approach retrieves this parameter from the RUN::CONFIG data bank, allowing the filter to adapt to varying experimental conditions and improving the accuracy of physics reconstruction. This update addressed maintenance challenges associated with static assumptions and aligned with ongoing project requirements. The work leveraged Java and applied skills in data analysis, physics simulation, and software engineering, resulting in more reliable filtering and improved data quality for downstream analyses.