What started as childhood curiosity, writing simple programs in MS-DOS and trying to understand how computers worked, eventually became a 19-year engineering career.

Today I work as a Senior Principal AI/ML Engineer, focused on real-time speech recognition systems running in production at scale. That means building and improving speech models, designing automated ML pipelines, and making sure these systems hold up where latency, robustness, and usability are non-negotiable.

One of my strengths is having worked across the full depth of this problem space. Before moving into AI and speech recognition, I spent years at the network layer, building systems that capture and process raw packet data with libpcap, and custom call recording solutions for speech analytics. That gave me hands-on fluency with VoIP signaling and media protocols: SIP, H.323, Alcatel NOE, RTP. It also pulled me naturally into real-time audio processing: noise reduction, silence detection, volume normalization. The quality of your input matters as much as the quality of your model.

On the ML side, I work primarily in Python and C++, training and optimizing ASR models with modern deep learning frameworks and end-to-end toolkits. The part I find most interesting is where research meets production: where an idea has to become something that actually runs, scales, and survives contact with real data.

I bring a research mindset to that work. I completed all doctoral coursework in a Computer Engineering PhD program with a perfect record, then chose to stay in industry, where I could apply that kind of analytical thinking directly to speech and language problems.

This blog is where I think out loud about AI and its impact, mostly on software engineering, but not exclusively. How engineering judgment gets valued, what changes when the tools start doing more of the work, what gets lost and what doesn't. Occasionally other things too, when something is worth writing about.

What still drives me is the same thing that started all of this: understanding how something works deeply enough to build something useful from it.