Worked on the django/django repository to optimize migration planning performance in large projects. Focused on improving the MigrationPlan generation process by replacing a list-based membership check with an OrderedSet, which reduced the computational complexity from quadratic to linear for large migration graphs. This change preserved the original insertion order and behavior while enhancing scalability and reducing planning time. The solution addressed a specific performance bottleneck in MigrationGraph._generate_plan and was validated in scenarios involving extensive migration dependencies. The work was implemented using Python and Django, demonstrating backend development skills and a collaborative approach to solving real-world performance challenges.