Research
Making simulation easier to question.
Our research direction is the intersection of discrete event simulation and AI: helping people explore alternatives and pursue goals with evidence they can inspect.
These are research questions, rather than claims of measured results or a released product.
Modeling
How can AI help describe a system?
Explore how natural language can help translate a process into entities, events, queues, and resources. A useful model still needs domain knowledge, explicit assumptions, and validation against the operation it represents.
Scenarios
Which alternatives are worth testing?
Investigate how AI can suggest changes to schedules, capacity, and operating rules. Each alternative should become a defined scenario that can be compared with a baseline.
Goals
How do we work backward from an outcome?
Study ways to connect a target with controllable inputs and practical constraints. Improving throughput may affect waiting time, utilization, and cost; those tradeoffs belong in the comparison.
Evidence
What makes a result useful?
Keep model versions, input data, random seeds, and run settings traceable. Repeated runs, confidence intervals, and sensitivity analysis help distinguish a promising change from a result that depends on one favorable assumption.