*3+ years relevant industrial / commercial experience or very relevant academic experience
*In-depth knowledge about probabilistic language modelling, including estimation methods, smoothing, pruning, efficient representation, interpolation, trigger-based models
*Experience in development and/or deployment of a speech recognition engines, with emphasis on efficient training techniques and large vocabulary systems
*Understanding of the technology and techniques used in ASR engines
*A good degree in an appropriate subject, PhD preferred
*Good software engineering skills, knowledge of scripting languages (e.g. shell scripting, Perl) and experience working under both Linux and Windows
Desirable
*Experience working with HTK, Sphinx and/or Julius
*Experience with probabilistic lattice parsing
*Have worked on a live application or service that includes speech recognition