At Deepgram, you will work alongside expert Research Scientists to design and train state-of-the-art deep learning algorithms for speech recognition, speaker classification, speech understanding, and natural language processing. With access to our massive GPU datacenters, you will be able to run fast-paced, iterative experiments and discuss results with the rest of the team, pushing the boundaries of accuracy in speech recognition.
Deepgram Interns are treated just like full time researchers, with significant freedom and ownership over their work. Interns take responsibility for outcome oriented projects and outcomes that our engineering, product, data operations team rely on.
What You'll Do
- Discover and create data suitable for training and validating deep learning models
- Identify state-of-the-art work and use it as a launching point for developing novel deep learning concepts
- Train and validate end-to-end deep learning models in the speech domain
- Carefully tune model hyperparameters to meet business accuracy and performance goals
You’ll Love This Role If You
- Are interested in working with and using large scale data to solve valuable automatic speech recognition problems.
- Enjoy training, validating, and testing machine learning models in common tensor frameworks, such as PyTorch
- Have a solution-oriented mindset and are comfortable using visualization tools (such as matplotlib or seaborn) to communicate findings and prove/disprove hypotheses
It’s Important To Us That You
- Are a Junior or Senior in a Undergrad Program or in a Graduate Program with a STEM background, preferably in fields such as Computer Science, Mathematics, Software Engineering, Physics, and Data Science.
- Comfortable using data and matrix libraries such as numpy and pandas, and using these libraries to load and manipulate large datasets
It Would Be Great if You Had
- Have experience with at least one programming language, preferably Rust or Python
- Taken graduate level courses in Machine Learning or Deep Learning