Yuvan Gorani

Machine Learning Intern @Stage

Noida, IN

At STAGE, I work on designing and developing a complete recommendation system that generates personalized movie and show suggestions for users based on their viewing behaviour. This end-to-end solution combines data engineering, machine learning, and backend API development. Key Highlights:1. Built a FastAPI backend to insert watch history, serve recommendations, and trigger updates.2. Implemented three recommendation models:•Statistical Model using a weighted rating formula.•Content-Based Filtering using genre similarity with cosine similarity and CountVectorizer.•Collaborative Filtering using the SVD algorithm from Surprise for personalized predictions.3. Designed a PostgreSQL schema to store user interactions, show/movie metadata, and generated recommendations daily, weekly and monthly.4. Introduced support for show-level aggregation, enabling show-based recommendations by grouping episodes using show_id and show_title.5. Added flexible filtering across models for:•content_type - show, movie, mixed•content_dialect - regional preferences (e.g, Har, Raj, Bho)•display_language - English, Hindi 6. Scheduled daily, weekly and monthly recommendation updates using a Docker-based cron job.7. Served all recommendations via API, including raw model outputs and final top-10 personalized blends.Focus Areas:Recommendation systems, backend development, data modeling, SQL, Python, FastAPI, Docker, machine learning (SVD, content similarity), and data pipelines.

New Digamber Public School

10th- 96.8% | 12th- 95.2%

Indian Institute of Technology, Kanpur

Bachelor of Technology - BTech