Fabián del Valle

About Fabián del Valle

Currently at J.P. Morgan Chase, I architect multi-agent AI orchestrators, production anomaly detection platforms at 100K+ portfolio scale, and LLM-powered pipelines that cut analyst workflows from hours to minutes. I\'ve led teams, shipped systems, and delivered ML that people actually depend on.My research background is in locality-sensitive hashing — I\'ve applied LSH to neural network pruning (99% FLOP reduction, 2.4x faster training) and scalable probabilistic clustering, work I did during my M.S. at NYU on a full-tuition scholarship. I think algorithmically, not just empirically.Before grad school I interned at AWS, Verizon, and Synopsys. ML Engineering, Applied Science, and AI/LLM Engineering.

Experience

1. Senior Associate Ai Engineer

   JPMorganChase

   Mar 2025 — Present · New York, NY, US

   Building production AI systems at one of the world\'s largest financial institutions.Designed and shipped a multi-agent AI orchestrator using Tool Registry, Agent Factory, and A2A messaging to coordinate distributed workflows across 10+ structured data sourcesArchitected a production anomaly detection platform monitoring 100K+ portfolios — reduced analyst review scope by 98% using LSH-based detection, cutting investigation time from 2 hours to 30 minutesExposed agentic workflows via FastAPI microservices, reducing data extraction turnaround from hours to minutesBuilt interpretable ML models (surrogate polynomial regression) to deliver feature-level attribution reports to 200+ stakeholdersLed ML integration into data pipelines, reducing downstream data defects by 80% and eliminating 10 manual validation roles.

Education

• University of Puerto Rico-Mayaguez

  Bachelor's Degree, Computer Science

• New York University

  Master's Degree, Computer Science