Local Shopping ML Pipeline

Overview

As a Software Engineering Intern at Google, I developed machine learning pipelines to enhance the Local Shopping team's ability to recommend relevant products to users based on proximity.

Technical Details

• Developed complete ML pipelines for data ingestion, transformation, training and evaluation using TensorFlow, TFX, Python, SQL and C++
• Experimented with models ranging from baseline heuristics and Linear Regression to Wide & Deep architectures
• Implemented feature transformation techniques using embeddings and spatial feature crosses to optimize location-based recommendations