Carlos Castillo

About Carlos Castillo

With more than 15 years of academic and industrial research experience, I build autonomous AI agents that reason, plan, and execute in complex environments. I specialize in training models for both visual analysis and agentic AI systems, translating research into production-ready solutions that deliver measurable impact.My work focuses on multi-agent orchestration, autonomous planning, and AI workflows that can navigate uncertainty while collaborating with humans. I leverage my computer vision expertise to create visually-intelligent agents that understand and interact with their environment. I also specialize in agent evaluation and trustworthy AI, ensuring reliable and safe deployment of autonomous systems.My expertise: Agentic AI systems, multi-agent frameworks, autonomous planning, agent evaluation, trustworthy AI, computer vision, model training, machine learning, deep learning, production AI deployment.

Experience

1. Principal Applied Scientist

   Amazon

   Aug 2022 — Present · Seattle, WA, US

   Building the next generation of autonomous AI agents that reason, plan, and execute complex tasks independently. I advance science and build production-ready systems that can navigate uncertainty, collaborate with humans, and deliver measurable business impact. From multi-agent orchestration to goal-driven AI workflows, I\'m architecting the intelligent systems that will define the future of work and decision-making.

Education

• University of Maryland

  Doctor of Philosophy (Ph.D.), Computer Science

  2005 — 2011

• Universidad Simón Bolívar

  Bachelor's degree, Computer Science

  1998 — 2003

Skills

• Matlab
• Image Processing
• Signal Processing
• Pattern Recognition
• Latex
• Experimental Design
• Algorithms
• Machine Learning
• Scientific Computing
• R
• Statistics
• Numerical Analysis
• Mathematical Modeling
• Computer Vision
• Experimentation
• Natural Language Processing