I’m a Senior Research Engineer with ~5 years of experience bridging research and production in AI systems. My current focus areas are multimodal search, agentic retrieval, and post-training methods for large language models.
At Bloomberg, I lead research into multimodal search architectures — evaluating Vision-Language Models against traditional text-based retrieval and implementing late-interaction techniques like ColPali-style scoring to improve retrieval precision across diverse data modalities. I also develop agentic retrieval pipelines that use query decomposition and recursive refinement to resolve complex, underspecified queries.
On the post-training side, I have hands-on experience with SFT, PeFT, and RL-based training methods, as well as designing LLM-as-a-judge evaluation workflows for assessing relevance and factuality in RAG systems.
I’m have contributed to open-source projects like LlamaIndex in the past, and have served as a reviewer for ACL ARR and ICAIF.
When I’m not coding, you’ll probably find me learning a new classical piano piece, experimenting with my home coffee setup, or scoping out new restaurants around NYC. I’ve got a growing list of travel destinations that I hope to check off soon.
Education
- M.Sc in Computer Science — University of Nebraska-Lincoln (4.0 GPA)
- B.Sc in Computer Science & Economics — University of Nebraska-Lincoln (3.91 GPA)
Location
- New York City, NY