Pranav Koundinya Amruthapuri

Built an end-to-end AI video generation platform combining playable games, AI-generated visuals, and automated transitions. Automated weekly creative generation for 240+ storylines across 3 regions and developed 4 reusable creative formats.

Improved LTV prediction for the GCC market by adapting country-level data and GCC-specific features using PySpark and SQL. Automated static ad creation and built a recovery and revenue analysis dashboard.

Automated inventory-tracking by implementing a dynamic-threshold based system, reducing manual monitoring time by 10 hours per week. Built an ML classification system to auto-categorize incoming stock, maintaining 88% accuracy.

Developed a U-Net CNN for automated road segmentation, achieving 91.17% test accuracy. Standardized and cleaned data, handled outliers, and fine-tuned hyperparameters to enhance model performance by 12%.

Grade: 9.61 CGPA

Pursued a Bachelor of Engineering in Artificial Intelligence and Machine Learning, achieving a CGPA of 9.61.

Completed the Machine Learning Specialization offered by Stanford Online and DeepLearning.AI.