Vyom Bhatt

Senior Data Scientist

• Job Details:Developed a customer lifetime value (CLV) model on Spark for precise margin predictions, optimizing marketing budgets, return policies, and contact center prioritization, resulting in $30M+ in savings. Led a team of four to build a call transcript labeling system using few-shot classification with RAG, improving accuracy by 42 percentage points over manual tagging and saving 10+ hours per month of manual effort. Built multiple product recommender engines, leveraging vision transformers for complementary product suggestions and neural collaborative filtering for personalized recommendations, boosting conversion rates by 2.4x. Engineered a data pipeline with 600+ features, integrating marketing, transaction, and browsing data, reducing data preparation time by 40% for ML models and customer analytics. Implemented multiple predictive models, including propensity models for site conversion, time-to-next purchase models, response models for campaign optimization, and sales forecasting models.

Data Scientist

• Job Details:Developed a YouTube channel recommender engine using NLP embeddings, improving channel performance by 22% over conventional ad targeting strategies. Designed a data lake integrating multiple geo-based data sources and built a regression model to identify high-potential geographies for effective ad targeting, securing 15+ clients and generating over $1M in revenue. Built a multi-touch attribution model using Shapley value and Markov chains to measure marketing channel impact, optimizing budget allocation and increasing ROI by 35%.