Rohan Mehta · Systems Engineer
Building models and systems for work that is difficult to specify, measure, and trust.
I’m a systems engineer with expertise in model-based systems engineering. I work across AI workflows, information retrieval, complex systems, and organizational knowledge.
Much of my work begins with processes that are qualitative, distributed across people and tools, and difficult to inspect. I work with subject-matter experts to turn that operational knowledge into explicit models, measurable pipelines, evaluation frameworks, and feedback loops.
I’m particularly interested in the infrastructure around AI: how organizations define what good performance means, preserve the evidence behind decisions, diagnose failure, and decide when a system has earned enough trust to act.
Trust does not come from choosing humans over machines. It comes from the institutional checks and balances around consequential decisions.
Current work
Chaos Coded Systems
As Principal Systems Architect, I translate operational knowledge into dependable data, automation, modeling, and AI-ready infrastructure.
Dynamical Systems Group
As a Research Engineer, I apply systems engineering to retrieval, agentic workflows, dataset design, evaluation, and simulation-driven research.
From signals to systems
I studied Physics and Electronics and Instrumentation Engineering at BITS Pilani. At TIFR’s FOTON Lab, I worked on time-domain simulations of terahertz metamaterial structures and co-authored an Optical Materials Express publication.
My thesis used machine learning to recover signals from noisy terahertz spectroscopy data. That work became an early foundation for the questions I still return to: what a model preserves, what it leaves out, and how we know when its output is reliable enough to use.
This site
I use this site to keep essays, research notes, finished projects, and work that is still taking shape.