I build speech and language systems end to end, from phoneme-level alignment through the service that deploys it. Linguistics before engineering, so the models get judged on whether they are fair to the speaker in front of them, not only on aggregate accuracy.
Most of my work is private or lives with the people I built it for. If something here is relevant to you, reach out and I am happy to walk you through it.
