I have been working on speech recognition-related problems for the last 15 years. I was a Staff Scientist at BBN, a Speech Scientist at Speechworks (later acquired by Nuance), a Core Technology Scientist at Scanscout (later acquired by Tremor Video).
While at CMU, I was one of the maintainers of the CMU Sphinx project. I mainly worked on Sphinx 3.X but later branched out on SphinxTrain. Later, I also worked on other components of the project.
I currently serve as a Principal Speech Architect at Voci Technology. My speciality is to maintain and further develop ASR-related software. I also heavily involved in other machine learning tasks such as emotion classification, gender detection as well as keyword spotting.
The Grand Janitor's Blog is the collection of my thoughts on speech recognition, machine learning and technology-related issues. Since 2016 April, I curated the Facebook Artificial Intelligence and Deep Learning (AIDL) Forum, we have an weekly office hour channel and a tied-in weekly newsletter.