Sotabase

Career

· Product, Anyscale2023–
· Engineering Manager, Anyscale2021–2023
· Team Lead, Anyscale2019–2021
· Graduate Student Researcher (PhD), Berkeley RISE Lab2017–2019
· Doctor of Philosophy - PhD Computer Science, University of California, Berkeley2017–2022
· Undergraduate Student Instructor, UC Berkeley2015–2019
· Software Engineering Intern, Zendesk2015–2015
· Research Assistant, UC Berkeley2014–2014
· Research Assistant, UCSF2014–2014

Publications (22)

Ray: A Distributed Framework for Emerging AI Applications
USENIX Symposium on Operating Systems Design and Implementation · 2017
1,516
cited
Tune: A Research Platform for Distributed Model Selection and Training
arXiv.org · 2018
1,056
cited
RLlib: Abstractions for Distributed Reinforcement Learning
International Conference on Machine Learning · 2017
981
cited
Ray RLLib: A Composable and Scalable Reinforcement Learning Library
Neural Information Processing Systems · 2017
178
cited
Benchmarks for reinforcement learning in mixed-autonomy traffic
Conference on Robot Learning · 2018
170
cited
SWIRL: A sequential windowed inverse reinforcement learning algorithm for robot tasks with delayed rewards
Int. J. Robotics Res. · 2018
74
cited
Real-Time Machine Learning: The Missing Pieces
USENIX Workshop on Hot Topics in Operating Systems · 2017
64
cited
HyperSched: Dynamic Resource Reallocation for Model Development on a Deadline
ACM Symposium on Cloud Computing · 2019
49
cited
HIRL: Hierarchical Inverse Reinforcement Learning for Long-Horizon Tasks with Delayed Rewards
arXiv.org · 2016
41
cited
RubberBand: cloud-based hyperparameter tuning
European Conference on Computer Systems · 2021
28
cited
Composing Meta-Policies for Autonomous Driving Using Hierarchical Deep Reinforcement Learning
arXiv.org · 2017
25
cited
Iterative Noise Injection for Scalable Imitation Learning
arXiv.org · 2017
22
cited
IMPACT: Importance Weighted Asynchronous Architectures with Clipped Target Networks
International Conference on Learning Representations · 2019
15
cited
Elastic Hyperparameter Tuning on the Cloud
ACM Symposium on Cloud Computing · 2021
7
cited
ESCHER: expressive scheduling with ephemeral resources
ACM Symposium on Cloud Computing · 2022
7
cited
SkyRL-Agent: Efficient RL Training for Multi-turn LLM Agent
arXiv.org · 2025
6
cited
REVEAL 2022: Reinforcement Learning-Based Recommender Systems at Scale
ACM Conference on Recommender Systems · 2022
2
cited
Ray RLlib: A Framework for Distributed Reinforcement Learning
2017
1
cited
Efficient Unicontact Grasping in Cluttered Scenes
2021
Examining Trends in Out-of-Domain Confidence
2021
Sotabase