Welcome! 🎉

As an electronics engineer with a penchant for music, my career has taken an uncommon trajectory. I’ve transitioned from being a sound recordist to a machine learning scientist/engineer. My curious and meticulous nature has led me to explore various domains deeply, for just wanting to know how things work. My journey started with carnatic classical voice training, eventually some rock and blues guitar, some classical violin, and have my ears set on improvization and jazz. I eventually got into music production and composition professionally, which brought me closer to signal processing and haptic feedback in music.

These diverse areas frequently involved adapting and iterating to improve audio outputs by optimizing a suitable statistical heuristic with gradual iterative steps, i.e. the doorstep of Machine Learning. Pretty soon, before anyone could say adaptive filter, I was poking around with hyperplanes, clusterings, regressions and got lost in the world of audio deep learning. I’ve spent time poking around with things like noise reduction, blind source separation, differentiable audio signal processing. More recently, my professional focus has been on speaker verification systems and leading the audio deepfake detection efforts at Resemble AI, while more generally, I spend my life overthinking the nuances of techno-feudalism with a side of artificial intelligence and the existential capitalist moloch traps that we’re seemingly, walking into constantly.

I’m always open to chat about these things or generally advise other humans that are finding themselves taking similar journies to mine, with the hope of re-cycling some my experiences into theirs.