Multimodal Learning For Campaign Classification
Mailchimp is the world's largest marketing automation platform. Over a billion emails are sent by it every day, which raises the question: what exactly are its users sending? We'll do a deep dive into the way that Mailchimp combines data across natural language and image modalities to generate numerical representations of email campaigns and make sense of users' content. Which allows it to empower small businesses by surfacing personalized recommendations.
Muhammed Ahmed is a Data Scientist at Mailchimp who specializes in natural language processing and deep learning. At Mailchimp, he has implemented several state-of-the-art deep learning models (ELMo, BERT, XLNet, RoBERTa, T5) for natural language understanding. His deployed models have been used for email campaign text classification tasks like spam detection and predicting user intent. Most recently, his focus has been on developing multimodal models which generate high quality numerical representations using both text and images.