Company Overview
Tomato.ai softens accents on calls. The company raised $12M and is led by 2 ex-Googlers who worked in the speech space for years. The founders previously sold 4 tech startups. The company is remote-first, based in the US, and hiring for this role in the US and world wide.
Pay range
Highly competitive compensation and benefits. Exact compensation may vary based on skills, experience, and location.
Location
Fully remote.
Responsibilities
- Build and scale the model inference infrastructure for real time streaming audio inputs to reduce latency and improve scalability & reliability.
- Develop tools to monitor the performance of inference infrastructure.
- Improve the CI/CD process.
- Develop and operate pipelines for large scale speech data processing using Apache Beam
- Develop algorithms for training data selection and augmentation for speech ML models
- Closely collaborate with the researchers for achieving the model performance goals.
Required Qualifications
- Hands-on experience in building large scale inference infrastructure.
- Good understanding of the state-of-the-art deep learning techniques.
- Proficient in Python and PyTorch
- Experience in Triton and TensorRT
- Effective communication skills.
- Ability to work independently in a remote-first environment.
Preferred Qualifications
- Experienced with audio streaming inputs.
- Familiarity with GCP.
- Proficient in C++
- Experience with speech or audio ML models is optional but a big plus
- Experience in site reliability engineering is helpful.