Sarah Gurev
Sarah Gurev is a machine-learning researcher who works on protein design and virology. Her EVEscape work uses artificial intelligence to predict viral evolutions and mutations, and a 2023 Nature paper with Nicole Thadani and Pascal Notin presents it. Trained on sequences from before 2020, EVEscape was as accurate as high-throughput experimental scans at anticipating pandemic variation for SARS-CoV-2. TIME included her in its 2024 TIME100 AI list. She earned an MS at MIT in 2023 and a PhD there in 2025, and she is now a post-doctoral fellow at FutureHouse.
- Now at
- FutureHouse
Timeline
-
2020
A computer-science degree at Stanford
In 2020, she completes a BS in computer science at Stanford.
-
2023
A master's degree at MIT
She earns an MS in electrical engineering and computer science at MIT in 2023.
-
2023
EVEscape published in Nature
She co-authors Learning from prepandemic data to forecast viral escape, a Nature paper from 2023, with Nicole Thadani and Pascal Notin.
-
2023
Forecasting pandemic variation
The paper reports that EVEscape, trained on sequences available before 2020, is as accurate as high-throughput experimental scans at anticipating pandemic variation for SARS-CoV-2.
-
2024
Named to TIME's 2024 AI list
TIME includes her in its 2024 TIME100 AI list. EVEScape, which she leads, uses artificial intelligence to predict viral evolutions and mutations.
-
2025
A PhD at MIT
In 2025, she completes a PhD in electrical engineering and computer science at MIT, working in the Debora Marks lab.
-
2026
A preprint on variant effects
In 2026, she co-authors a bioRxiv preprint, Evaluating variant effect prediction across viruses, with Noor Youssef and Debora Marks.
-
As of October 2026
A FutureHouse postdoctoral fellowship
She is a post-doctoral fellow at FutureHouse, working as a machine learning researcher on protein design and virology. Her research covers deep generative models of viral and immune protein sequences and structures.
Sources and credits
Photo: We have no published photo of Sarah Gurev yet. Send us a free-licensed photo.