Automated Speech Recognition (ASR) Engineer - XM Discover
The Qualtrics XM Platform™ is a system of action that helps businesses to attract customers who stay longer and buy more, to engage and empower employees to do the best work of their lives, to develop breakthrough products people love, and to build a brand people can’t imagine living without.
Joining Qualtrics means becoming part of a team bold enough to chase breakthrough experiences - like building a technology that will be a force for good. A team committed to diversity, equity, and inclusion because of a conviction that every voice holds value, with a vision for representation that matches the world around us and inclusion that far exceeds it. You could belong to a team whose values center on transparency, being all in, having customer obsession, acting as one team, and operating with scrappiness. All so you can do the best work of your career.
We believe every interaction is an opportunity. Are we yours?
About the role:
As an ASR Engineer on the Speech Engineering team, you will be working with software developers and implementation consultants to evaluate and improve the accuracy of Qualtrics' automated speech recognition (ASR) systems. This role involves a blend of research, innovation, and ingenuity in order to solve very real and practical ASR problems. In this role you will be doing things such as:
- Building acoustic and domain models
- Performing WER analysis as well as Qualtrics -specific accuracy analysis
- Researching applicable noise reduction techniques
- Collaborating with subject matter experts to tune ASR output in support of specific business problems
- Assembling training data to support tuning activities
- Defining best practices for handling audio quality problems and other ASR-related issues
About the team:
The Qualtrics Speech Engineering team is a deeply technical team that works alongside software engineers to get the most out of the Qualtrics ASR system. The team uses state of the art tools and software to evaluate and improve automated transcription processes, as well as to implement new ASR techniques and enhance existing approaches. The team's research helps guide the development of software and models that are a part of the automated transcription process, while the team's practical focus ensures that all work can be used in real-world applications.
About you:
You are a curious, enthusiastic, and technical candidate with a proven record in audio processing or digital audio manipulation. A good combination of the following skills and experiences makes you a good fit for this role:
- Degree in Linguistics, Speech Pathology, Computational Linguistics, or Audio Engineering
- Experience managing or manipulation audio files for human voice and speech
- Knowledge of Digital Audio Containers, Codecs, and Compression algorithms (G.711, G.729, PCM, etc)
- Facility with Linux command line tools, for example ffmpeg and sox
- Familiar with machine learning and speech recognition