Aleksandra Faust
Aleksandra Faust is a Serbian-American computer scientist, AI researcher, and technology executive. She is the Chief AI Officer at Genesis Molecular AI, having previously served as a Research Director at Google DeepMind, and a Principal Investigator at Sandia National Laboratories.
Faust is recognized for establishing principles of AI-driven scalable autonomy, particularly in the field of Automated Reinforcement Learning. Her research focuses on treating the entire system design pipeline as a learnable, sequential decision-making problem—an approach she has applied to scalable autonomy in robotics, generative AI, and drug discovery. Contributions include the "Pearl" biomolecular foundation model, the self-improvement training methods used in Google's Gemini models, and the "Levels of AGI" framework. In 2020, she received the IEEE Early Career Award in Robotics and Automation.
Education
Faust received her Bachelor of Science in Mathematics and Computer Science from the University of Belgrade. She earned a Master of Science in Computer Science from the University of Illinois at Urbana-Champaign in 2004. In 2014, Faust completed her Ph.D. in Computer Science at the University of New Mexico under the supervision of Lydia Tapia. Her dissertation, "Reinforcement Learning and Planning for Preference Balancing Tasks," won the Tom L. Popejoy Dissertation Prize, the university's highest dissertation honor.Career
Faust was a Senior R&D Engineer at Sandia National Laboratories. She subsequently joined Waymo in 2015, focusing on machine learning for motion planning.In 2017, Faust joined Google Brain, eventually rising to Director of Research at Google DeepMind, where she led scalable autonomy and reinforcement learning research.
In June 2025, Faust was appointed Chief AI Officer of Genesis Molecular AI. In October 2025, she and her team released the technical report for the "Pearl" foundation model for atomic placement in biomolecular structures, reportedly the first model that outperforms AlphaFold 3.
Automated Reinforcement Learning (AutoRL)
Faust co-authored the paper that founded Automated Reinforcement Learning, a term her research is credited with coining. AutoRL automates the design of the learning agents themselves. She co-authored the field's first survey, and served as the Program Chair for the AutoML conference in 2023.Sustainable Training Methodologies
A central tenet of Faust's work is the reliance on accessible, imperfect data to overcome scarcity in high-stakes fields. Her research in robotics, web agents, and drug discovery utilizes synthetic, simulated, and noisy data to propel progress where expert demonstrations are rare or nonexistent.Robotics and Motion Planning
In robotics, Faust bridges the gap between sensing, motion planning, and control using machine learning. She created "PRM-RL," a method that combines sampling-based planning with reinforcement learning to enable long-range autonomous navigation, winning the Best Paper in Service Robotics award at ICRA 2018.Faust was also an early advocate for generalist robot models capable of navigating diverse physical spaces without retraining. She established the theoretical foundations for this generalization as well as self-supervised methods for a learning-based robotics stack without computationally expensive methods. She later expanded this approach to hardware-software co-design, characterizing dependencies between sensors, compute, and machine learning models. This interdisciplinary work earned the Best of IEEE Computer Architecture Letters runner-up award and an IEEE Micro Top Picks Honorable Mention. Her contributions to the field were recognized with the IEEE Early Career Award in Robotics and Automation in 2020.
Generative AI and Autonomous Agents
Faust led the development of Web Agents, recognized as the first fully autonomous, open-ended task agents on the web. This technology was integrated into Google Assistant. To measure industry progress, Faust co-authored "Levels of AGI," a framework operationalizing the path to artificial general intelligence. The framework has been discussed in media outlets including Bloomberg News, The Economist, and Forbes.Awards and honors
- Fellow of the IEEE, 2026 "for contributions to technical leadership in scalable learning-based autonomy and foundation models"
- IEEE Micro Top Picks Honorable Mention
- 50 Women in Robotics you need to know about, Women in Robotics
- Best Paper of IEEE Computer Architecture Letters runner-up
- IEEE Early Career Award in Robotics and Automation
- ICRA Best Paper in Service Robotics
- Distinguished Alumna, University of New Mexico School of Engineering
- Tom L. Popejoy Dissertation Prize Winner, University of New Mexico