Chandan V.

1. Content Moderation: Designing algorithm to flog out NSFW content before it gets uploaded onto platform by fine-tuning OPEN Al CLIP algorithms Achievements: Reducing manual efforts by almost 70% with a 0.5% chances of NSFW content getting missed by the algorithm Scaling the architecture to handle 50 thousand videos per day 2. Search: Developing a search system to help find users relevant content they are looking for 3. Personalized Recommendations: Developing a personalized recommendation system for the discover page on the app which enriches user experience

1. News Meta Tagging: Tagging in the incoming news content (videos and text} into various categories using Natural language processing and speech—to-text based technologies Achievements: Increased the CTR by 13% by generating accurate tags and categories which in turn provided better recommendations 2. Video genre classification: Classification of Zee5 premium video contents into pre-defined genres using Video analytics, computer vision, and deep learning Achievements: Successfully implemented research paper published by IBM for optimal sequential grouping for scene detection