Vince Castillo Ph.D.
Logistics professor exploring how generative and agentic AI reshape supply chain decisions, work, and education.
Research
I study how AI systems behave as economic and operational actors — and what that means for the organizations that deploy them.
- Measuring how AI systems negotiate, allocate, forecast, and fail.
- Designing agentic systems for supply chain decision-making.
- Comparing AI and human actors through behavioral experiments.
- Google Scholar
Publications and citations.
- Do Humans and GAI See Eye to Eye? Implications of LLM Scoring Volatility in Supplier Evaluations
Journal of Business Logistics · AI and human procurement evaluations.
- Validating Generative Agent-Based Models for Logistics and Supply Chain Management Research
Preprint · Comparing simulated AI behavior with human behavior.
- Hybrid Fleets
- Agent Farm Live site
Project explanation and code, alongside the live deployment.
- AI in Supply Chain (AiSC)
My lab’s work at the intersection of AI and supply chain management.
Teaching
Bringing AI into supply chain education through learning tools, classroom simulations, and executive education.
- AI Fluency Masterclass for Supply Chain
Executive education at Ohio State’s Fisher College of Business.
- Supply Chain Brutus
A supply chain learning tool.
- Supply Chain Wars
A supply chain simulation.
- Teaching with an AI-built classroom simulation
Featured video · See the simulation in a teaching context.
Connect
I’m interested in industry collaborations that bring real operational problems to agentic AI research.
Contact me about speaking engagements.
- OSU profile
My faculty profile at Fisher College of Business.
- LinkedIn
I occasionally post about AI and supply chain management.
- GitHub
Code and projects.
- Hugging Face
My Hugging Face profile.
Under the hood
My research program centers on machine behavior — the empirical study of how AI systems act as decision-makers in economic and operational settings. That means treating LLMs not as tools to be prompted, but as agents to be measured: how do they negotiate, allocate, forecast, and fail? And what happens when you put them inside supply chain workflows that were designed for humans?
- LLM benchmarking in logistics and operations contexts
- Agentic AI system design for supply chain decision-making
- Behavioral experiments comparing AI and human actors
- Applied generative AI in procurement, warehousing, and last-mile
I'm actively seeking industry collaborators for agentic AI research — particularly organizations willing to bring real operational problems to the table. If that's you: [email protected].