I am an AI Research Scientist at MSL FAIR at Meta in Seattle. At FAIR, my research has spanned topics in NLP such as model architectures, evaluation, and data collection. I’ve contributed to efficient and multimodal architectures—including the Byte Latent Transformer (ACL 2025 Outstanding Paper) and Chameleon. My evaluation work spans assessing the validity of AI benchmarks, dynamic data collection and benchmarking through Dynatask/Dynabench, and applying item response theory to NLP leaderboards. I currently work on AI & Human Co-Improvement.
I completed my PhD in Computer Science at the University of Maryland at College Park titled “Evaluating Machine Intelligence with Question Answering” in 2021 where I worked in the Computational Linguistics and Information Processing (CLIP) Lab with Jordan Boyd-Graber.
I am also a nature/wildlife/landscape photographer, avid skier, climber, hiker, and violinist.
Contact
Contact me at me@pedro.ai. Recruiters, please read this page before contacting me.
Industry Experience
- I have been an AI Research Scientist at Meta since 2021 and in FAIR since 2022.
- Throughout my undergraduate and graduate studies I interned extensively and accumulated over three years of industry experience.
- I was a research intern at Google AI-Zürich working on question answering and fact verification during Winter 2019/20.
- At Facebook Conversational AI I worked on conversational information seeking which resulted in an EMNLP 2020 publication.
- During summer 2018 I was a research intern at Microsoft Research in the Information and Data Science Group.
- I was a data science intern at Riot Games and Oracle/Datalogix.
- Prior to starting my PhD, I was a data scientist at Zillow Group/Trulia.
Education
I defended my PhD dissertation in computer science at University of Maryland at College Park in April 2021. I earned my Bachelor’s Degree in Computer Science from the University of California at Berkeley in December 2014. During my undergraduate degree, I was a research assistant in UC Berkeley’s AMPLab, UC Berkeley’s Astronomy Department, and Boise State University’s Cryosphere Group.
Service
- Area Chair for COLING 2022
- Organizer for the Dynamic Adversarial Data Collection (DADC) Workshop at NAACL 2022
- Session Chair at NAACL 2022
- Grant Proposal Reviewer for Meta AI RFP: Dynabench Data Collection and Benchmarking Platform
- Reviewer for the Workshop on Insights from Negative Results in NLP
- Reviewer for ACL ARR 2022
- Reviewer for ACL ARR 2021
- Reviewer for EMNLP 2021
- Sub-committee Co-Chair for Diversity and Inclusion for ACL 2021
- Reviewer for ACL 2021
- Reviewer for NAACL 2021
- Secondary Reviewer for EMNLP 2020
- Secondary Reviewer for ACL 2019
- Founding member and former captain of the Colorado Data Science Team.
Certifications and Professional Organizations
- Member of Association for Computational Linguistics
- Certified avalanche instructor for the National Ski Patrol
- Professional member of the American Avalanche Association.
Avalanche Education
As a bilingual avalanche safety instructor, I have taught hundreds of students throughout Idaho, Oregon, California, Colorado, Argentina, and Chile how to safely enjoy backcountry skiing and riding. In my blog, I’ve posted several of my lectures and written on snowgeek.org as well. Fun fact: I got my nickname “Pinguino” while teaching in Argentina (e.g., the favicon for my site is a penguin).