At IBM I worked on a collaborative team to develop a protocol for designing, simulating, deploying, and characterizing numerically-optimized gates on IBM quantum processors. A central contribution of mine was developing a systematic calibration procedure to optimize control waveforms and steer experimental gate performance toward intended targets, with randomized benchmarking used to validate the results. To demonstrate the practical utility of this calibration routine, I designed and executed a proof-of-concept experiment generating gates robust against spectator errors between nearest-neighbor qubits – a problem with no analytical solution where the value of custom numerical design becomes evident. I regularly utilized python/qiskit (qiskit pulse, qiskit experiment, qiskit dynamics) and Git version control, in addition to low-level IBM hardware languages. This work resulted in a first-author manuscript (arXiv:2412.05230) and taught me how to identify compelling demonstration opportunities that showcase novel capabilities while addressing genuine technical challenges in quantum systems.