CONVERTLY | Data Scientist (dataset preparation and model evaluation)
Defined model evaluation frameworks to assess scoring accuracy and recommendation quality using real business engagement data. Prepared datasets suitable for model training by cleaning and analyzing raw performance inputs, including missing-value handling and metric normalization. Built processing and pipeline architecture in Python to support the optimization layer for recommendations. • Cleaned and analyzed social media performance data for training readiness. • Normalized engagement metrics and handled missing values. • Engineered Python model pipeline architecture for optimization intelligence. • Evaluated recommendation performance against engagement outcomes.