Designing complex systems,
from first principles

I'm drawn to problems that are messy, multi-layered, and slightly underdefined, the kind where you have to model the system before you can even begin solving it. I'm a Computer Engineering student at Thapar, focused on backend systems, distributed architectures, and AI infrastructure. My current interests sit at the intersection of AI systems and research: agentic architectures, interpretability, and how large models reason, generalize, and break. I'm also building evaluation frameworks for LLM behavior at openfolie, an applied AI safety research collective, where the interesting questions aren't benchmarks, they're about what "correct behavior" even means when the system is agentic.