Ameya’s innovative research proposal earned her a spot among 24 Etsy Summer of Vision Fellows from across the country. During the early days of the COVID-19 pandemic, when fabric masks had not yet made their debut on Etsy, she collaborated closely with a senior machine learning engineer and a senior data scientist to develop a proposal aimed at revolutionizing the e-commerce experience.

​

Her project initially focused on classifying fabric patterns using advanced AI techniques. By applying transfer learning and hyperparameter tuning on convolutional neural networks (CNNs) implemented in TensorFlow, Ameya leveraged a MobileNetV2 model to classify a dataset of 15,000 Etsy fabric images across 25 different patterns. Recognizing the potential to enhance product search further, she expanded the scope of her research to explore the classification of patterns on fabric mask listings, a timely project given the evolving market during the pandemic.

​

Moreover, her research delved into fine-tuning a model that could accurately identify patterns by visual features. This precision was crucial for filtering out incorrect search results across various product listings including masks made of multiple kinds of fabrics. The overarching goal was to create a more accurate, personalized e-commerce experience by refining search results based on color, pattern, and texture, ultimately allowing Etsy customers to find exactly what they envisioned.

 

The culmination of her work was presented to fellow Summer of Vision Fellows, mentorship leaders, and Etsy’s Director of Engineering & Applied Machine Learning, with the intent of moving the proposal toward implementation. This initiative not only highlighted her technical prowess but also her forward thinking approach to integrating AI with real-world e-commerce challenges.

​​​