Deep Learning Engineer
At AssemblyAI, we use State-of-the-Art Deep Learning to build the #1 most accurate Speech-to-Text API for developers. We’re backed by leading investors in Silicon Valley like Y Combinator, John and Patrick Collison (Stripe), Nat Friedman (GitHub), and Daniel Gross.
Customers use our API to transcribe phone calls, meetings, videos, podcasts, and other types of media. Our accurate transcripts are used to power features like visual voicemail, call analytics, closed captioning, meeting summaries, and a slew of other features.
We deploy our Deep Learning models into production to process millions of API requests per day.
About the Role
We are growing rapidly and looking for an experienced Deep Learning Engineer to join our Speech Recognition Team, with a focus on large, distributed model training. We’re a small, creative, and democratic team interested in pushing the state of the art forward. You’ll be leading the efforts on things like:
- Scaling up distributed training to train models with 100s of millions of parameters across dozens of V100 GPUs
- Optimizing distributed training to cut training time of extremely large models down from weeks to days
- Researching and implementing techniques to make large models more efficient for inference; such as model pruning, quantization, and model distillation
- Researching and implementing techniques for model ensembling
- 3-5+ years of experience with Python
- 2+ years of experience with CUDA
- 2+ years of experience with Deep Learning frameworks like PyTorch and TensorFlow
- 2+ years of experience training distributed deep learning models on GPUs
- Some experience with modern Deep Learning based ASR systems (CTC, LAS, RNNTs)
Discover More AI Jobs:
- Address United States