I have twenty years experience in constructing production software in intelligent/adaptive systems. My corporate research focus for the last decade has been primarily on text classification and automatic language analysis. Design and development of a fast real-time web content classification API at RuleSpace stands out as one of my most consumed implementations powering filtering products in several of the largest telecom/carriers and search companies worldwide effectively filtering billions of requests.
My academic focus is on modeling continuous speech. Specifically, my work entails constructing data-derived coarticulation models for continuous speech to allow speaker-specific speech synthesis, formant tracking, making conversational speech more intelligible, characterizing dysarthria, increasing accuracy in automatic speech recognition, and increasing our understanding of human speech production.
[AI/machine learning skills]: genetic algorithms/programming, Bayesian classification, maximum entropy classification, clustering, fuzzy logic, support vector machines (SVM), data mining, statistics, intelligent system architecture/design
[Language/API skills]: c, python, numpy, bash, UNIX, and library/API design/implementation.
[Text specialties]: topical crawling, corpus building, content filtering, processing "big data", feature discovery/engineering, and text classification in both real-time and offline systems.
[Speech specialties]: spectrogram reading, DSP algorithms, intelligibility/perceptual testing, formant trajectory behavior and modeling, coarticulation, acoustic-phonetics, speech corpus project management/development and experimental design.
[Technology Management]: entrepreneurship, mentoring and project management in applied scientific research.