Maryam Fazel-Zarandi is a Research Engineering Manager at Facebook AI Research (FAIR), leading the US east coast NLP, Speech and Core ML researchers. Her research interests are in the areas of natural language understanding and dialog systems and her current research focuses on reasoning in large language models and self-supervised learning for speech. She has over 10 years of industry experience and extensive experience in collaborating with engineering teams to deliver science solutions to production. Prior to joining FAIR in 2021, she was an Applied Science Manager at Amazon working on Conversational AI for Alexa, where she supported the development of Alexa Conversations and led a team of scientists and engineers working on data-driven dialog management, incremental learning in dialog systems, and reasoning for language understanding and dialog management. Before joining Amazon in 2017, she was a Senior Research Scientist at Nuance Communications, working on open-domain natural language understanding and dialog systems and question answering. Maryam received her Ph.D. in Computer Science from University of Toronto in 2013.
2021 - Present: Research Engineering Manager, Meta AI - Facebook AI Research (FAIR)
2017 - 2021: Applied Science Manager and Senior Machine Learning Scientist, Amazon Alexa
2013 - 2017: Senior Research Scientist, Nuance Communications
2007 - 2013: PhD, Department of Computer Science, University of Toronto
2005 - 2007: MSc, Department of Computer Science, University of Toronto
2001 - 2005: BASc, Department of Computer Engineering, Sharif University of Technology
- Our paper Scaling Speech Technology to 1,000+ Languages is now on arXiv. Checkout the blog post and demo!
- Our paper Evolutionary-scale prediction of atomic-level protein structure with a language model was accepted to Science!
- Our paper Cocktail HuBERT: Generalized Self-Supervised Pre-training for Mixture and Single-Source Speech was accepted to ICASSP 2023
- Our paper ROSCOE: A Suite of Metrics for Scoring Step-by-Step Reasoning was accepted to ICLR 2023
- Checkout our work on Evolutionary-scale prediction of atomic level protein structure with a language model
- Our paper Towards Large-Scale Interpretable Knowledge Graph Reasoning for Dialogue Systems was accepted to Findings of ACL 2022
- Our paper on Logical Reasoning for Task Oriented Dialogue Systems was accepted to ECNLP 5 @ ACL
- Our paper on Alexa Conversations was accepted to NAACL-HLT 2021
- Our paper on Dialog Simulation with Realistic Variations for Training Goal-Oriented Conversational Systems was accepted to the Human in the Loop Dialogue Systems Workshop at NeurIPS 2020
- Learn more about the science behind the Alexa Conversations
- Our paper on Data Augmentation for Training Dialog Models Robust to Speech Recognition Errors was accepted to the NLP for Conversational AI workshop at ACL 2020
- Talk: Amazon Workshop at ICASSP 2020
I will be presenting our work at the 3rd Workshop on Conversational AI at NeurIPS 2019 in Vancouver: