National Science Foundation Ph.D Fellow at Caltech, advised by Yisong Yue.

I do research in the field of Reinforcement Learning, with a primary focus on developing methods to guarantee desirable post-learning performance of computational agents.

Outside of research, I run a consulting firm. We deliver machine learning and data-driven solutions to companies of all sizes with a special focus on recommendation systems, prediction, and computer vision domains. We love helping companies achieve their KPIs. If your company needs help in their machine learning goals, reach out to me.

Previously, I worked as a ML Engineer and Researcher at several companies. In a previous life, I spent some time in the finance world working in Venture Capital, Investment Banking and Prop Trading.

Outside of work, I love trying new hobbies. Some recent ones include starting a tango orchestra and learning to sail. My current favorite cocktail: Last Word

Interested in chatting about Reinforcement Learning, engineering or new interests to explore?
Contact: cvoloshin [at] caltech [dot] edu


My research can be broadly categorized into two subfields of RL:

Off-Policy Policy Evaluation (OPE)

OPE is the task of predicting an agent's performance before deployment, the last step of the testing pipeline. This line of work is meant to ensure that the agent will behave in a desirable way.

Constrained Policy Learning

Unlike OPE, ensuring desirable performance can also be baked into the learning process itself by requiring the agent to satisfy a particular set of constraints. We study the ways to both express and enforce the constraints during learning.

  • Evaluation
  • Learning

Education + WorkPDF


Ph.D in Computing & Mathematical Sciences

2019 - 2023 Expected

California Institute of Technology, Pasadena, CA

B.S in Applied & Computational Mathematics

2013 - 2017

California Institute of Technology, Pasadena, CA

Current Work

Graduate Student

2019 - Present

Caltech, Pasadena, CA

  • I carry out research in Reinforcement learning with a primary focus in Off-policy Evaluation and Policy Optimization under constraints

Research Intern

2022 - 2022

Argo AI, Pittsburgh, PA

  • Reinforcement Learning and Off-policy Evaluation in the Autonomous Vehicle space.