At Sanyu Infotech, I cleaned and analyzed a dataset of 50,000+ records to prepare it for modeling. I built and evaluated machine learning models using Python and TensorFlow.
I exposed model results through Flask REST APIs for real-time predictions. I also independently developed and deployed PopcornPicks, a movie recommendation system using embedding similarity and TMDB API data.
For my Retail Sales & Profitability BI Dashboard project, I built a MySQL star-schema warehouse to analyze retail orders. I used SQL transformations and window functions to prepare reporting data, then designed an interactive Power BI dashboard.
On SilentVerify, I collected and analyzed simulated user-behavior sessions for a passive CAPTCHA detection system. The model achieved 92% classification accuracy and reduced user friction by 75%.
