Employer: CaptionCall, LLC

Job Title: Speech Scientist II

Worksite address: 4192 S. Riverboat Road, Salt Lake City, UT 84123

Position fully remote. Work from home benefit available.

Job Duties:
Design and conduct experiments to find methods to improve speech recognition accuracy, including data collection and preparation, running recognition software, and analyzing results using Python languages, Linux, Bash, and PRAAT.
Perform experimental design, data analysis, statistics, and data science.
Develop and simulate machine learning strategies for using multiple sources of evidence to predict and improve speech recognition quality.
Train and test neural networks, maximum entropy classifiers, and other machine learning methods.
Design, build, and evaluate neural networks using PyTorch and TensorFlow neural network tools.
Build statistical estimation and classification models using Scikit-learn.
Write software to integrate speech and language code into a real-time system.
Build acoustic and language models for speech recognition, using Kaldi and wav2vec2 speech recognition software.
Benchmark speech recognition accuracy of experimental recognizers compared to commercial systems.
Develop methods for estimating speech recognition confidence, including methods that operate independently of the speech recognition system.
Implement and validate techniques for adding vocabulary words to speech recognizers and modifying probability weights for vocabulary words to be emphasized or suppressed for hybrid and end-to-end speech recognizers.
Implement and evaluate new methods for speech recognition by training and evaluating end-to-end neural networks by experimenting with alternative network topologies and training strategies.
Curate speech recordings and transcripts for use in training speech recognizers by developing methods for automatically correcting errors, detecting missing data, and converting the data format for use by machine learning tools.
Research, design, develop, and productize AI technology for new markets by applying machine learning to building AI-based services.
This is a fully remote position. Work from home benefit offered. Worksites are unanticipated.

Degree Requirement: Bachelor of Science Degree in Computer Science, Software Engineering, or closely related field.

Experience required: 1 year of experience as a Software Engineer or an Engineer in an IT field.

Skill set Requirements (Special Requirements):
1 year of experience as a Software Engineer or an Engineer in an IT field must include 6 months in a Speech Recognition or Machine Learning field.

Demonstrable knowledge of and prior experience in the following:
Python languages, Linux, Bash, and PRAAT.
Kaldi speech recognition software.
Speech recognition theory and methods.
PyTorch and TensorFlow neural network tools.
Scikit-learn machine learning library.
Statistical modeling for estimation and classification.
Training and evaluating neural networks.
Applying machine learning to building AI-based services.

Travel Required: none

Work Hours: 8:00 am. – 5:00 pm, Monday – Friday.

To Apply: Please submit an application and upload a resume on our website (link removed) Only qualified applicants need apply. Do not apply if you are an H-1 or F-1 visa holder, or if you otherwise require employer sponsorship to obtain U.S. employment authorization. Do not apply if you are not eligible to be employed in the U.S.