Summary
I am a Research Assistant in Dr. Van Durme’s lab at Johns Hopkins University, where I work on NLP, argument mining, and legal reasoning. I earned my M.S. in Engineering Management from Johns Hopkins in 2025, under the guidance of Dr. Mahyar Fazlyab. Before this, I led teams in hospitality, retail, software, and sports.
My research focuses on interpretable, safe, and reliable language models for legal analysis. I use declarative language models, case grammar, and LLMs to produce evidence-grounded outputs with calibrated uncertainty. I aim to support legal reasoning by surfacing relevant facts, constructing and critiquing arguments, and aiding high-stakes decisions. My current work examines how models can make evidence evaluation more transparent and capture the different ways humans weigh legal factors.
- Applied ML/NLP
- Argument mining
- IR/RAG
- Legal reasoning
Full details live in my CV. Thank you for visiting!
Research
- Chain-of-Syllogisms: Unifying Analysis & Conclusions Boosts Argument Mining — Paper · Paper
- Mining Legal Arguments in U.S. Corporate Law — Submitted to ARR · Code · Paper
- AAO non-precedent dataset + meta-interpreter for decision inference — Ongoing · Code · Presentation
Education
- M.S., Engineering Management (Smart Devices/ECE Track) — Johns Hopkins University (2025)
- B.Eng., Industrial Engineering — Universidad de Lima (2018)
- Graduate Program, Corporate Law — ESAN Business School (2017)
Projects
Chain-of-Syllogisms
New argument scheme for passage classification + Corporate law dataset
Nmbr9
Digital version of NMBR 9 with a custom RL agent (SAM).
Mentat
Medical‑AI platform work: annotation workflows plus ML/NLP features (information extraction, generation, rewriting).
Warehouse Layout Optimization
Fast heuristic for optimizing large-scale warehouse design using PRM and the Capacity Scaling Algorithm. Closed form solution.
Adaptive Beamforming
Ported ST MEMS-mic beamforming libraries to STM32L476 + X-NUCLEO-CCA02M2.