Software Engineer - Machine Learning Infrastructure
- San Francisco, CA, United States
Whisper exists to improve the human senses. Our initial product is the world’s first noise cancelling hearing aid system: it analyzes your audio in real time, automatically filtering out noise and amplifying the sounds you want to hear. Unlike traditional hearing aids, which simply amplify everything in the room, Whisper amplifies the person you’re listening to based on millions of audio prints it learns over time so it’s able to pick out who you’re listening to, even in the noisiest restaurant. Based in San Francisco, Whisper is lucky to have the support of great investors including Sequoia Capital, First Round Capital, LUX Ventures, and more. Whisper is an an equal opportunity employer committed to a diverse workforce with an inclusive working environment for everyone to do their best work. We do not discriminate on the basis of race, ethnicity, religion, gender, gender identity, sexual orientation, age, marital status, veteran status, or disability status.
THE ROLE Building state-of-the-art deep learning acoustic models that improve people’s listening experience is the centerpiece of the Whisper hearing product, and great infrastructure and tools is what enables us to iterate quickly and develop new technology in this area. Machine learning infrastructure spans everything from creating new data pipelines to ingest an ever-growing, proprietary real-world acoustic dataset to building model quantization pipelines that will take new Tensorflow models and have them run efficiently on our embedded platform. This role is perfect for someone who wants to learn about AI and has a penchant for building new, low-maintenance solutions that make a company move faster.
RESPONSIBILITIES ● Build, maintain and improve new systems to handle user feedback, error tracking, distributed data analysis, model training, and other core machine learning tasks. ● Maintain and improve on various continuous integration systems that improve machine learning models, especially on hardware. ● Develop new internal tools to increase experimentation velocity on new models (e.g. to help make it easier to understand results). ● Improve performance across a wide number of systems, including data generation, model training, experimentation, and more. ● Take ownership of data management at Whisper from a security and privacy point of view, with anything from debug logs to sensitive user data being stored.
Whisper exists to improve the human senses. Our hearing aids uses deep learning to remove background noise so you can hear clearly in a restaurant or at home when the dishwasher is running – this is the number one complaint in the industry, and our recent results reduce unwanted noise by 10x any other brand.
Skip straight to final-round interviews by applying through Triplebyte.