Aug 2023 – Present
Senior Member of Technical Staff, Cohere
Post-training LLM agents to do useful work for enterprises.
I am a Senior Member of Technical Staff at Cohere, where I work on reinforcement learning and synthetic data for post-training code agents.
Before Cohere I was a research engineer at Reality Defender. Prior to that I did machine learning research in the Machine Learning for Language (ML²) group at New York University, where I had the privilege of being advised by Kyunghyun Cho and Sam Bowman, and working closely with Cheolhyoung Lee and Jason Phang.
Experience
Mar 2023 – Aug 2023
Research Engineer, Reality Defender
Detection models for synthetic and manipulated media.
Nov 2021 – Mar 2023
Researcher, ML² Group, New York University
Research on neural language models, advised by Kyunghyun Cho and Sam Bowman.
Research
Frontier models
Command A: An Enterprise-Ready Large Language Model
Technical report, 2025
tl;dr a technical report detailing the development of Cohere's flagship model, Command A.
Self-improvement
Self-Improving Robust Preference Optimization
ICLR 2025
tl;dr SRPO is a robust direct alignment method that unlocks inference-time self-refinement.
Policy optimization
Contrastive Policy Gradient: Aligning LLMs on sequence-level scores in a supervised-friendly fashion
EMNLP 2024
tl;dr CoPG is a scalable, principled off-policy RL algorithm that estimates the optimal policy without importance sampling.
Language modeling
A Non-monotonic Self-terminating Language Model
ICLR 2023
tl;dr prevents non-terminating sequences that arise from incomplete decoding algorithms.
Teaching
May 2022
Deep Learning for NLP — African Masters of Machine Intelligence
Teaching assistant and lab instructor for a modern NLP foundations course taught by Kyunghyun Cho and Duygu Ataman.
Spring 2022
DS-UA 203: Machine Learning for Language Understanding — New York University
Section leader and teaching assistant for a research-oriented natural language understanding course taught by Sam Bowman.
Education
2023
2022
B.A. Computer Science, minor in Mathematics
Courant Institute, New York University
CeDiD 22LH-QC4F-EFIF
Projects
Dec 2021
A Survey of Transfer Learning in Uncertainty Aware Computer-Aided Diagnosis
CSCI-GA 2271: Computer Vision, Rob Fergus
tl;dr transfer learning for uncertainty-aware brain and kidney tumor segmentation.