The work is centered on debugging and system-level behavior. That includes board bring-up, tracing hardware/software interactions, and isolating issues that don’t show up cleanly—timing problems, power instability, inconsistent sensor data, or communication failures across interfaces like SPI, I2C, UART, or RF links. In one case, I redesigned an embedded platform to reduce power consumption by 50% while also shrinking PCB area by 26%. That required changes across both hardware and firmware—power architecture, component selection, and how the system actually behaved over time, not just in short tests. I also work on improving sensing performance in noisy environments. Using MATLAB and signal processing techniques, I’ve helped refine detection algorithms for seismic systems, focusing on improving accuracy without increasing power or compute requirements. Across projects, a consistent pattern is working at the boundary between layers—where firmware meets hardware, or where system assumptions start to break. The approach is usually iterative: measure what the system is actually doing, narrow down failure modes, and make targeted changes that hold up outside the bench environment. Some teams bring me in early to think through architecture and tradeoffs. Others bring me in when something isn’t behaving and needs to be understood quickly. In both cases, the goal is the same—build systems that are stable, efficient, and predictable once they’re in use.